Trabajo Final Maestria UdeSA_ Código DRKG.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "77zD4aqad7ao",
   "metadata": {
    "id": "77zD4aqad7ao"
   },
   "outputs": [],
   "source": [
    "#Importación de librerías básicas útiles (asumiendo que se tiene instalado Pytorch):\n",
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import random\n",
    "import torch"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aed2df81",
   "metadata": {
    "id": "aed2df81"
   },
   "source": [
    "# **Contenido de los archivos.tsv:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3f7eb208",
   "metadata": {
    "id": "3f7eb208"
   },
   "outputs": [],
   "source": [
    "#drkg.tsv contiene el DRKG original en el formato de tripletes (h, r, t). Se carga en un DataFrame:\n",
    "df_drkg = pd.read_csv('drkg.tsv', sep='\\t', header=None) #Especificar la carpeta donde está guardado el archivo!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c628d6b2",
   "metadata": {
    "id": "c628d6b2"
   },
   "source": [
    "De cada base de datos fuente, se extrajo una lista de tripletes en la forma(head-entity, relation-type, tail-entity). Para representar las entidades, se usó un identificador de tipo de entidad seguido de un ID único de la entidad específica, por ejemplo, Gene::22947. Para representar relaciones, se usó una convención que combina el nombre de la fuente de datos, el nombre de la relación y los tipos de entidades head y tail que están involucradas, por ejemplo DGIDB::INHIBITOR::Gene:Compound. El orden de las entidades h y t es de suma importancia."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0f198313",
   "metadata": {
    "id": "0f198313"
   },
   "outputs": [],
   "source": [
    "#Se renombran las columnas del dataset:\n",
    "df_drkg.columns = ['Head Entity','Relation','Tail Entity']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a15dd9ad",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1729192833572,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "a15dd9ad",
    "outputId": "7ff79207-52f9-46b4-bf09-9641cd34ffef",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Head Entity                        Relation  Tail Entity\n",
      "0  Gene::2157  bioarx::HumGenHumGen:Gene:Gene   Gene::2157\n",
      "1  Gene::2157  bioarx::HumGenHumGen:Gene:Gene   Gene::5264\n",
      "2  Gene::2157  bioarx::HumGenHumGen:Gene:Gene   Gene::2158\n",
      "3  Gene::2157  bioarx::HumGenHumGen:Gene:Gene   Gene::3309\n",
      "4  Gene::2157  bioarx::HumGenHumGen:Gene:Gene  Gene::28912\n",
      "5  Gene::2157  bioarx::HumGenHumGen:Gene:Gene    Gene::811\n",
      "6  Gene::2157  bioarx::HumGenHumGen:Gene:Gene   Gene::2159\n",
      "7  Gene::2157  bioarx::HumGenHumGen:Gene:Gene    Gene::821\n",
      "8  Gene::2157  bioarx::HumGenHumGen:Gene:Gene   Gene::5627\n",
      "9  Gene::2157  bioarx::HumGenHumGen:Gene:Gene   Gene::5624\n"
     ]
    }
   ],
   "source": [
    "print(df_drkg.head(10)) #Primeros 10 registros del dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bqpp_6XzgB2k",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1729192833573,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "bqpp_6XzgB2k",
    "outputId": "f858860c-2b90-477c-de07-4bebd585198d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5874261\n"
     ]
    }
   ],
   "source": [
    "print(len(df_drkg)) #Cantidad de tripletes alamcenados en el dataset (tamaño del Grafo de Conocimiento o Knowledge Graph)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "26e705bc",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 363
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1729192833573,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "26e705bc",
    "outputId": "f2d659a9-3b2a-4a34-e0e2-798cbdce4015"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"df_drkg\",\n  \"rows\": 10,\n  \"fields\": [\n    {\n      \"column\": \"Head Entity\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 10,\n        \"samples\": [\n          \"Gene::2624\",\n          \"Gene::1630\",\n          \"Gene::1385\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Relation\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 8,\n        \"samples\": [\n          \"Hetionet::GiG::Gene:Gene\",\n          \"Hetionet::GpBP::Gene:Biological Process\",\n          \"DRUGBANK::ddi-interactor-in::Compound:Compound\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Tail Entity\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 10,\n        \"samples\": [\n          \"Compound::DB00928\",\n          \"Gene::9423\",\n          \"Pathway::PC7_7457\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-14bfdf2d-6d5b-4a39-8b21-04f4e0ad9a69\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Head Entity</th>\n",
       "      <th>Relation</th>\n",
       "      <th>Tail Entity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1292440</th>\n",
       "      <td>Compound::DB08894</td>\n",
       "      <td>DRUGBANK::ddi-interactor-in::Compound:Compound</td>\n",
       "      <td>Compound::DB11127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2454646</th>\n",
       "      <td>Gene::1630</td>\n",
       "      <td>Hetionet::GiG::Gene:Gene</td>\n",
       "      <td>Gene::9423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3262234</th>\n",
       "      <td>Gene::2961</td>\n",
       "      <td>Hetionet::Gr&gt;G::Gene:Gene</td>\n",
       "      <td>Gene::1398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5525296</th>\n",
       "      <td>Gene::54821</td>\n",
       "      <td>STRING::BINDING::Gene:Gene</td>\n",
       "      <td>Gene::5516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>869004</th>\n",
       "      <td>Compound::DB01046</td>\n",
       "      <td>DRUGBANK::ddi-interactor-in::Compound:Compound</td>\n",
       "      <td>Compound::DB13025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3433628</th>\n",
       "      <td>Gene::1385</td>\n",
       "      <td>Hetionet::GpPW::Gene:Pathway</td>\n",
       "      <td>Pathway::PC7_7457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2204392</th>\n",
       "      <td>Gene::493</td>\n",
       "      <td>Hetionet::GpBP::Gene:Biological Process</td>\n",
       "      <td>Biological Process::GO:0098655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3806073</th>\n",
       "      <td>Anatomy::UBERON:0002107</td>\n",
       "      <td>Hetionet::AeG::Anatomy:Gene</td>\n",
       "      <td>Gene::5160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86554</th>\n",
       "      <td>Gene::2624</td>\n",
       "      <td>DGIDB::OTHER::Gene:Compound</td>\n",
       "      <td>Compound::DB00928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2260760</th>\n",
       "      <td>Gene::51451</td>\n",
       "      <td>Hetionet::GpBP::Gene:Biological Process</td>\n",
       "      <td>Biological Process::GO:1903362</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-14bfdf2d-6d5b-4a39-8b21-04f4e0ad9a69')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-14bfdf2d-6d5b-4a39-8b21-04f4e0ad9a69 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-14bfdf2d-6d5b-4a39-8b21-04f4e0ad9a69');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-cfa6daa6-1799-476a-be47-9a0ab8649ada\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-cfa6daa6-1799-476a-be47-9a0ab8649ada')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-cfa6daa6-1799-476a-be47-9a0ab8649ada button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                     Head Entity  \\\n",
       "1292440        Compound::DB08894   \n",
       "2454646               Gene::1630   \n",
       "3262234               Gene::2961   \n",
       "5525296              Gene::54821   \n",
       "869004         Compound::DB01046   \n",
       "3433628               Gene::1385   \n",
       "2204392                Gene::493   \n",
       "3806073  Anatomy::UBERON:0002107   \n",
       "86554                 Gene::2624   \n",
       "2260760              Gene::51451   \n",
       "\n",
       "                                               Relation  \\\n",
       "1292440  DRUGBANK::ddi-interactor-in::Compound:Compound   \n",
       "2454646                        Hetionet::GiG::Gene:Gene   \n",
       "3262234                       Hetionet::Gr>G::Gene:Gene   \n",
       "5525296                      STRING::BINDING::Gene:Gene   \n",
       "869004   DRUGBANK::ddi-interactor-in::Compound:Compound   \n",
       "3433628                    Hetionet::GpPW::Gene:Pathway   \n",
       "2204392         Hetionet::GpBP::Gene:Biological Process   \n",
       "3806073                     Hetionet::AeG::Anatomy:Gene   \n",
       "86554                       DGIDB::OTHER::Gene:Compound   \n",
       "2260760         Hetionet::GpBP::Gene:Biological Process   \n",
       "\n",
       "                            Tail Entity  \n",
       "1292440               Compound::DB11127  \n",
       "2454646                      Gene::9423  \n",
       "3262234                      Gene::1398  \n",
       "5525296                      Gene::5516  \n",
       "869004                Compound::DB13025  \n",
       "3433628               Pathway::PC7_7457  \n",
       "2204392  Biological Process::GO:0098655  \n",
       "3806073                      Gene::5160  \n",
       "86554                 Compound::DB00928  \n",
       "2260760  Biological Process::GO:1903362  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Mostar 10 registros al azar:\n",
    "df_drkg.sample(n=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e2744f2",
   "metadata": {
    "id": "6e2744f2"
   },
   "outputs": [],
   "source": [
    "#entity2src.tsv contiene un mapeo de entidades en el DRKG a sus fuentes originales de donde fueron extraídas (Este paso no influye en la aplicación de reposicionamiento de drogas):\n",
    "columnas = ['Entity_name','Source1','Source2','Source3','Source4','Source5','Source6','Source7','Source8']\n",
    "df_entity2src = pd.read_csv('entity2src.tsv', sep='\\t', names = columnas, header = None, dtype = str) #Especificar la carpeta donde está guardado el archivo!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dLErZduN7L3b",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1729192834842,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "dLErZduN7L3b",
    "outputId": "6ae52df1-22c7-4fe8-eb18-fd0bfb98c9da"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97238, 9)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_entity2src.shape #Dimensiones del DataFrame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "39da512a",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 502
    },
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1729192834842,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "39da512a",
    "outputId": "64645150-17da-4a94-a03c-4a7fdad4ed0a",
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "repr_error": "0",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-fae7e5ad-873e-4409-8609-682e0eb4b7da\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Source1</th>\n",
       "      <th>Source2</th>\n",
       "      <th>Source3</th>\n",
       "      <th>Source4</th>\n",
       "      <th>Source5</th>\n",
       "      <th>Source6</th>\n",
       "      <th>Source7</th>\n",
       "      <th>Source8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23911</th>\n",
       "      <td>Gene::5314</td>\n",
       "      <td>[Hetionet] Biomedical knowledge graph https://...</td>\n",
       "      <td>[STRING] https://string-db.org/</td>\n",
       "      <td>[GNBR] Data extracted from biomedical texts se...</td>\n",
       "      <td>[IntAct] https://www.ebi.ac.uk/intact/</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80189</th>\n",
       "      <td>Biological Process::GO:0051450</td>\n",
       "      <td>[Hetionet] Biomedical knowledge graph https://...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91700</th>\n",
       "      <td>Side Effect::C0005411</td>\n",
       "      <td>[Hetionet] Biomedical knowledge graph https://...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46606</th>\n",
       "      <td>Compound::DB02975</td>\n",
       "      <td>[Drugbank] https://www.drugbank.ca/atc</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73431</th>\n",
       "      <td>Biological Process::GO:0002415</td>\n",
       "      <td>[Hetionet] Biomedical knowledge graph https://...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46831</th>\n",
       "      <td>Compound::DB03233</td>\n",
       "      <td>[Drugbank] https://www.drugbank.ca/atc</td>\n",
       "      <td>[Bibliograpic sources] Collection of datasourc...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35848</th>\n",
       "      <td>Gene::922</td>\n",
       "      <td>[Hetionet] Biomedical knowledge graph https://...</td>\n",
       "      <td>[STRING] https://string-db.org/</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88920</th>\n",
       "      <td>Cellular Component::GO:0031264</td>\n",
       "      <td>[Hetionet] Biomedical knowledge graph https://...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66374</th>\n",
       "      <td>Atc::N03AG05</td>\n",
       "      <td>[Drugbank] https://www.drugbank.ca/atc</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87813</th>\n",
       "      <td>Molecular Function::GO:0070733</td>\n",
       "      <td>[Hetionet] Biomedical knowledge graph https://...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-fae7e5ad-873e-4409-8609-682e0eb4b7da')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-fae7e5ad-873e-4409-8609-682e0eb4b7da button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-fae7e5ad-873e-4409-8609-682e0eb4b7da');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-d16f58ba-f279-4346-847a-516db5162978\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-d16f58ba-f279-4346-847a-516db5162978')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-d16f58ba-f279-4346-847a-516db5162978 button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                          Entity_name  \\\n",
       "23911                      Gene::5314   \n",
       "80189  Biological Process::GO:0051450   \n",
       "91700           Side Effect::C0005411   \n",
       "46606               Compound::DB02975   \n",
       "73431  Biological Process::GO:0002415   \n",
       "46831               Compound::DB03233   \n",
       "35848                       Gene::922   \n",
       "88920  Cellular Component::GO:0031264   \n",
       "66374                    Atc::N03AG05   \n",
       "87813  Molecular Function::GO:0070733   \n",
       "\n",
       "                                                 Source1  \\\n",
       "23911  [Hetionet] Biomedical knowledge graph https://...   \n",
       "80189  [Hetionet] Biomedical knowledge graph https://...   \n",
       "91700  [Hetionet] Biomedical knowledge graph https://...   \n",
       "46606             [Drugbank] https://www.drugbank.ca/atc   \n",
       "73431  [Hetionet] Biomedical knowledge graph https://...   \n",
       "46831             [Drugbank] https://www.drugbank.ca/atc   \n",
       "35848  [Hetionet] Biomedical knowledge graph https://...   \n",
       "88920  [Hetionet] Biomedical knowledge graph https://...   \n",
       "66374             [Drugbank] https://www.drugbank.ca/atc   \n",
       "87813  [Hetionet] Biomedical knowledge graph https://...   \n",
       "\n",
       "                                                 Source2  \\\n",
       "23911                    [STRING] https://string-db.org/   \n",
       "80189                                                NaN   \n",
       "91700                                                NaN   \n",
       "46606                                                NaN   \n",
       "73431                                                NaN   \n",
       "46831  [Bibliograpic sources] Collection of datasourc...   \n",
       "35848                    [STRING] https://string-db.org/   \n",
       "88920                                                NaN   \n",
       "66374                                                NaN   \n",
       "87813                                                NaN   \n",
       "\n",
       "                                                 Source3  \\\n",
       "23911  [GNBR] Data extracted from biomedical texts se...   \n",
       "80189                                                NaN   \n",
       "91700                                                NaN   \n",
       "46606                                                NaN   \n",
       "73431                                                NaN   \n",
       "46831                                                NaN   \n",
       "35848                                                NaN   \n",
       "88920                                                NaN   \n",
       "66374                                                NaN   \n",
       "87813                                                NaN   \n",
       "\n",
       "                                      Source4 Source5 Source6 Source7 Source8  \n",
       "23911  [IntAct] https://www.ebi.ac.uk/intact/     NaN     NaN     NaN     NaN  \n",
       "80189                                     NaN     NaN     NaN     NaN     NaN  \n",
       "91700                                     NaN     NaN     NaN     NaN     NaN  \n",
       "46606                                     NaN     NaN     NaN     NaN     NaN  \n",
       "73431                                     NaN     NaN     NaN     NaN     NaN  \n",
       "46831                                     NaN     NaN     NaN     NaN     NaN  \n",
       "35848                                     NaN     NaN     NaN     NaN     NaN  \n",
       "88920                                     NaN     NaN     NaN     NaN     NaN  \n",
       "66374                                     NaN     NaN     NaN     NaN     NaN  \n",
       "87813                                     NaN     NaN     NaN     NaN     NaN  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Selección de 10 registros aleatoriamente:\n",
    "df_entity2src.sample(n=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3rIrcFs9bHfY",
   "metadata": {
    "id": "3rIrcFs9bHfY"
   },
   "source": [
    "Se observa que existen entidades que aparecen en más de una base de datos!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b772b7de",
   "metadata": {
    "id": "b772b7de"
   },
   "outputs": [],
   "source": [
    "#relation_glossary.tsv, contiene un glosario de las relaciones en el DRKG, más información asociada con sus fuentes, si están disponibles (Paso que no influye en la aplicación de reposicionamiento de drogas):\n",
    "df_rel_glos = pd.read_csv('relation_glossary.tsv', sep='\\t', dtype = str) #Especificar la carpeta donde está guardado el archivo!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bf33ab8b",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 450
    },
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1729192836974,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "bf33ab8b",
    "outputId": "64976c5c-ee03-4c84-d516-6bfaa355afaf"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"df_rel_glos\",\n  \"rows\": 10,\n  \"fields\": [\n    {\n      \"column\": \"Relation-name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 10,\n        \"samples\": [\n          \"DRUGBANK::treats::Compound:Disease\",\n          \"INTACT::ADP RIBOSYLATION REACTION::Gene:Gene\",\n          \"GNBR::A+::Compound:Gene\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Data-source\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 7,\n        \"samples\": [\n          \"STRING\",\n          \"IntAct\",\n          \"GNBR\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Connected entity-types\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          \"Compound:Gene\",\n          \"Compound:Disease\",\n          \"Gene:Gene\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Interaction-type\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 9,\n        \"samples\": [\n          \"antibody\",\n          \"ADP ribosylation reaction\",\n          \"agonism, activation\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Description\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 5,\n        \"samples\": [\n          \"This is a label given by the reporting source to an interaction that doesn't belong to other interaction types, as defined by the reporting source.\",\n          \"interaction that doesn't belong to other interaction types.\",\n          \"An antibody interaction occurs when an antibody drug specifically binds the target molecule.\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Reference for the description\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"http://www.dgidb.org/getting_started\",\n          \"https://www.ebi.ac.uk/ols/ontologies/mi/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FMI_0557\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-5d38f680-4db6-48e5-9428-599cb19f0a04\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Relation-name</th>\n",
       "      <th>Data-source</th>\n",
       "      <th>Connected entity-types</th>\n",
       "      <th>Interaction-type</th>\n",
       "      <th>Description</th>\n",
       "      <th>Reference for the description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>STRING::BINDING::Gene:Gene</td>\n",
       "      <td>STRING</td>\n",
       "      <td>Gene:Gene</td>\n",
       "      <td>binding</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>INTACT::ADP RIBOSYLATION REACTION::Gene:Gene</td>\n",
       "      <td>IntAct</td>\n",
       "      <td>Gene:Gene</td>\n",
       "      <td>ADP ribosylation reaction</td>\n",
       "      <td>Involves the addition of one or more ADP-ribos...</td>\n",
       "      <td>https://www.ebi.ac.uk/ols/ontologies/mi/terms?...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>DGIDB::OTHER::Gene:Compound</td>\n",
       "      <td>DGIDB</td>\n",
       "      <td>Compound:Gene</td>\n",
       "      <td>other</td>\n",
       "      <td>This is a label given by the reporting source ...</td>\n",
       "      <td>http://www.dgidb.org/getting_started</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>bioarx::DrugVirGen:Compound:Gene</td>\n",
       "      <td>BIBLIOGRAPHY</td>\n",
       "      <td>Compound:Gene</td>\n",
       "      <td>interaction</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>Hetionet::GpBP::Gene:Biological Process</td>\n",
       "      <td>HETIONET</td>\n",
       "      <td>Biological Process:Gene</td>\n",
       "      <td>participation</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>GNBR::A+::Compound:Gene</td>\n",
       "      <td>GNBR</td>\n",
       "      <td>Compound:Gene</td>\n",
       "      <td>agonism, activation</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>GNBR::E::Compound:Gene</td>\n",
       "      <td>GNBR</td>\n",
       "      <td>Compound:Gene</td>\n",
       "      <td>affects expression/production (neutral)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>DGIDB::ANTIBODY::Gene:Compound</td>\n",
       "      <td>DGIDB</td>\n",
       "      <td>Compound:Gene</td>\n",
       "      <td>antibody</td>\n",
       "      <td>An antibody interaction occurs when an antibod...</td>\n",
       "      <td>http://www.dgidb.org/getting_started</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>DRUGBANK::treats::Compound:Disease</td>\n",
       "      <td>DRUGBANK</td>\n",
       "      <td>Compound:Disease</td>\n",
       "      <td>Compound treats the disease</td>\n",
       "      <td>We extract  the  associated  conditions  relat...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>STRING::OTHER::Gene:Gene</td>\n",
       "      <td>STRING</td>\n",
       "      <td>Gene:Gene</td>\n",
       "      <td>other</td>\n",
       "      <td>interaction that doesn't belong to other inter...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5d38f680-4db6-48e5-9428-599cb19f0a04')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-5d38f680-4db6-48e5-9428-599cb19f0a04 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-5d38f680-4db6-48e5-9428-599cb19f0a04');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-32b25dd5-70f9-4203-b5d0-acb73760fb1b\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-32b25dd5-70f9-4203-b5d0-acb73760fb1b')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-32b25dd5-70f9-4203-b5d0-acb73760fb1b button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                                    Relation-name   Data-source  \\\n",
       "94                     STRING::BINDING::Gene:Gene        STRING   \n",
       "80   INTACT::ADP RIBOSYLATION REACTION::Gene:Gene        IntAct   \n",
       "10                    DGIDB::OTHER::Gene:Compound         DGIDB   \n",
       "104              bioarx::DrugVirGen:Compound:Gene  BIBLIOGRAPHY   \n",
       "74        Hetionet::GpBP::Gene:Biological Process      HETIONET   \n",
       "19                        GNBR::A+::Compound:Gene          GNBR   \n",
       "28                         GNBR::E::Compound:Gene          GNBR   \n",
       "4                  DGIDB::ANTIBODY::Gene:Compound         DGIDB   \n",
       "18             DRUGBANK::treats::Compound:Disease      DRUGBANK   \n",
       "98                       STRING::OTHER::Gene:Gene        STRING   \n",
       "\n",
       "      Connected entity-types                         Interaction-type  \\\n",
       "94                 Gene:Gene                                  binding   \n",
       "80                 Gene:Gene                ADP ribosylation reaction   \n",
       "10             Compound:Gene                                    other   \n",
       "104            Compound:Gene                              interaction   \n",
       "74   Biological Process:Gene                            participation   \n",
       "19             Compound:Gene                      agonism, activation   \n",
       "28             Compound:Gene  affects expression/production (neutral)   \n",
       "4              Compound:Gene                                 antibody   \n",
       "18          Compound:Disease              Compound treats the disease   \n",
       "98                 Gene:Gene                                    other   \n",
       "\n",
       "                                           Description  \\\n",
       "94                                                 NaN   \n",
       "80   Involves the addition of one or more ADP-ribos...   \n",
       "10   This is a label given by the reporting source ...   \n",
       "104                                                NaN   \n",
       "74                                                 NaN   \n",
       "19                                                 NaN   \n",
       "28                                                 NaN   \n",
       "4    An antibody interaction occurs when an antibod...   \n",
       "18   We extract  the  associated  conditions  relat...   \n",
       "98   interaction that doesn't belong to other inter...   \n",
       "\n",
       "                         Reference for the description  \n",
       "94                                                 NaN  \n",
       "80   https://www.ebi.ac.uk/ols/ontologies/mi/terms?...  \n",
       "10                http://www.dgidb.org/getting_started  \n",
       "104                                                NaN  \n",
       "74                                                 NaN  \n",
       "19                                                 NaN  \n",
       "28                                                 NaN  \n",
       "4                 http://www.dgidb.org/getting_started  \n",
       "18                                                 NaN  \n",
       "98                                                 NaN  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Selección de 10 registros aleatoriamente:\n",
    "df_rel_glos.sample(n=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "CzCwYlzEczSg",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 36
    },
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1729192838453,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "CzCwYlzEczSg",
    "outputId": "f7e1b3af-34ce-404a-a681-da7d53361e7c"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "type": "string"
      },
      "text/plain": [
       "'In inhibitor interactions, the drug binds to a target and decreases its expression or activity. Most interactions of this class are enzyme inhibitors, which bind an enzyme to reduce enzyme activity.'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Veo el campo de descripción de un registro específico:\n",
    "df_rel_glos['Description'].iloc[8]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60d1dafd",
   "metadata": {
    "id": "60d1dafd"
   },
   "source": [
    "# **Embeddings moleculares:**\n",
    "Obtenidos a partir de Graph Neural Networks preentrenadas mediante la combinación de predicción supervisada de propiedades moleculares y enfoques de aprendizaje autosupervisado.\n",
    "Se determinaron 4 variantes de embeddings moleculares usando DGL-LifeSci (DGL-LifeSci es un paquete de Python para la aplicación de redes neuronales de grafos a varias tareas en química y biología."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3ab384db",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 3202,
     "status": "ok",
     "timestamp": 1729192844119,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "3ab384db",
    "outputId": "e64cd6a5-189b-4b95-a42e-e2d41c361130"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8807, 300)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 1) Embeddings obtenidos a partir de un modelo preentrenado para predecir estructuras de grafos alrededor de subgrafos moleculares:\n",
    "embed_contextpred = np.load('./embed/mol_contextpred.npy') #Especificar la carpeta donde está guardado el archivo!\n",
    "embed_contextpred.shape #8807 embeddings de dimensión 300"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bb736a1f",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 637,
     "status": "ok",
     "timestamp": 1729192844754,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "bb736a1f",
    "outputId": "42b5ca7b-aa45-44ee-f310-44ea28bd2cb8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8807, 300)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 2) Embeddings obtenidos a partir de un modelo preentrenado para maximizar la información mutua entre representaciones locales de nodos y una representación global del grafo:\n",
    "embed_infomax = np.load('./embed/mol_infomax.npy') #Especificar la carpeta donde está guardado el archivo!\n",
    "embed_infomax.shape #8807 embeddings de dimensión 300"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0d7f9f27",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 3367,
     "status": "ok",
     "timestamp": 1729192848119,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "0d7f9f27",
    "outputId": "83531a80-1c28-42cb-89e6-411e2fba7df0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8807, 300)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 3) Embeddings obtenidos a partir de un modelo preentrenado para promover a nodos vecinos a tener representaciones similares y forzar a nodos dispares a tener representaciones diferentes:\n",
    "embed_edgepred = np.load('./embed/mol_edgepred.npy') #Especificar la carpeta donde está guardado el archivo!\n",
    "embed_edgepred.shape #8807 embeddings de dimensión 300"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bb5642c4",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1932,
     "status": "ok",
     "timestamp": 1729192850049,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "bb5642c4",
    "outputId": "48bf84b5-346e-4b79-b4ef-d3300d3d253c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8807, 300)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 4) Embeddings obtenidos a partir de un modelo preentrenado para predecir atributos enmascarados aleatoriamente de nodos y aristas:\n",
    "embed_masking = np.load('./embed/mol_masking.npy') #Especificar la carpeta donde está guardado el archivo!\n",
    "embed_masking.shape #8807 embeddings de dimensión 300"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6398ada2",
   "metadata": {
    "id": "6398ada2"
   },
   "source": [
    "# Análisis de calidad del DRKG"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef9eb018",
   "metadata": {
    "id": "ef9eb018"
   },
   "source": [
    "El análisis de calidad del DRKG se realiza en base a un análisis de similaridad entre los diferentes tipos de relaciones existentes en el grafo basado en el solapamiento de nodos y aristas entre tipos de relaciones. Específicamente, se determinan los coeficientes de Jaccard para nodos y aristas así como también un coeficiente de solapamiento de aristas entre tipos de estas.\n",
    "Este análisis nos permite evaluar la calidad del grafo construido. Para más información al respecto, ver el paper original en https://github.com/gnn4dr/DRKG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "IW2d855p-MFU",
   "metadata": {
    "id": "IW2d855p-MFU"
   },
   "outputs": [],
   "source": [
    "triplets = df_drkg.values.tolist() #Generamos una lista de tripletes a partir del dataframe inicial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "AQj-zqvfZIEH",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1729192854341,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "AQj-zqvfZIEH",
    "outputId": "08640d13-bdfd-4209-b1ba-5a335dca8c2d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Gene::2157', 'bioarx::HumGenHumGen:Gene:Gene', 'Gene::811']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "triplets[5] # Cada elemento de la lista obtenida es una lista definiendo un triplete [Head Entity, Relation, Tail Entity] dado."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Blzcg11s5HyG",
   "metadata": {
    "id": "Blzcg11s5HyG"
   },
   "outputs": [],
   "source": [
    "#Esta porción de código encuentra entidades únicas según su tipo. Genera un diccionario con key igual a un tipo de entidad y values igual a todas las\n",
    "#entidades de ese tipo, sin repertirlas.\n",
    "entity_dictionary={}\n",
    "def insert_entry(entry,ent_type,dic):\n",
    "    if ent_type not in dic:\n",
    "        dic[ent_type]={}\n",
    "    ent_n_id=len(dic[ent_type])\n",
    "    if entry not in dic[ent_type]:\n",
    "         dic[ent_type][entry]=ent_n_id\n",
    "    return dic\n",
    "\n",
    "for triple in triplets:\n",
    "    src = triple[0]\n",
    "    split_src=src.split('::')\n",
    "    src_type=split_src[0]\n",
    "    dest = triple[2]\n",
    "    split_dest=dest.split('::')\n",
    "    dest_type=split_dest[0]\n",
    "    insert_entry(src,src_type,entity_dictionary)\n",
    "    insert_entry(dest,dest_type,entity_dictionary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "vPzfJBvS_Rum",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1729024248652,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "vPzfJBvS_Rum",
    "outputId": "aa13a0f3-5e22-4877-988f-85b2df71e92c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cellular Component\n"
     ]
    }
   ],
   "source": [
    "entity_types_list = list(entity_dictionary.keys())\n",
    "print(random.choice(entity_types_list)) #Imprime un tipo de entidad al azar."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "jWQ96_tSZ4lK",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 6,
     "status": "ok",
     "timestamp": 1729024248652,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "jWQ96_tSZ4lK",
    "outputId": "c850d030-3237-4142-bb15-d8fe00d74518"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Cellular Component::GO:0005913': 0, 'Cellular Component::GO:0043025': 1, 'Cellular Component::GO:0005874': 2, 'Cellular Component::GO:0031253': 3, 'Cellular Component::GO:0035686': 4, 'Cellular Component::GO:0031080': 5, 'Cellular Component::GO:1990204': 6, 'Cellular Component::GO:0000974': 7, 'Cellular Component::GO:0071556': 8, 'Cellular Component::GO:0097422': 9, 'Cellular Component::GO:0005871': 10, 'Cellular Component::GO:0031252': 11, 'Cellular Component::GO:0000930': 12, 'Cellular Component::GO:0005938': 13, 'Cellular Component::GO:0032039': 14, 'Cellular Component::GO:0031227': 15, 'Cellular Component::GO:0008180': 16, 'Cellular Component::GO:0005770': 17, 'Cellular Component::GO:0044297': 18, 'Cellular Component::GO:0034680': 19, 'Cellular Component::GO:0098791': 20, 'Cellular Component::GO:0043073': 21, 'Cellular Component::GO:0005773': 22, 'Cellular Component::GO:0000159': 23, 'Cellular Component::GO:0043564': 24, 'Cellular Component::GO:0034399': 25, 'Cellular Component::GO:0008287': 26, 'Cellular Component::GO:0045111': 27, 'Cellular Component::GO:0036477': 28, 'Cellular Component::GO:0042175': 29, 'Cellular Component::GO:0048471': 30, 'Cellular Component::GO:0031966': 31, 'Cellular Component::GO:0044291': 32, 'Cellular Component::GO:0045177': 33, 'Cellular Component::GO:1990351': 34, 'Cellular Component::GO:0043005': 35, 'Cellular Component::GO:0098857': 36, 'Cellular Component::GO:0055029': 37, 'Cellular Component::GO:0099512': 38, 'Cellular Component::GO:0045121': 39, 'Cellular Component::GO:0005743': 40, 'Cellular Component::GO:0099503': 41, 'Cellular Component::GO:0000776': 42, 'Cellular Component::GO:0001726': 43, 'Cellular Component::GO:0044815': 44, 'Cellular Component::GO:0019898': 45, 'Cellular Component::GO:0030667': 46, 'Cellular Component::GO:0061695': 47, 'Cellular Component::GO:0030315': 48, 'Cellular Component::GO:0043235': 49, 'Cellular Component::GO:0016604': 50, 'Cellular Component::GO:0000790': 51, 'Cellular Component::GO:0005912': 52, 'Cellular Component::GO:0009986': 53, 'Cellular Component::GO:0001750': 54, 'Cellular Component::GO:0005758': 55, 'Cellular Component::GO:0031306': 56, 'Cellular Component::GO:0001533': 57, 'Cellular Component::GO:0043204': 58, 'Cellular Component::GO:1990904': 59, 'Cellular Component::GO:0005789': 60, 'Cellular Component::GO:0098793': 61, 'Cellular Component::GO:0030014': 62, 'Cellular Component::GO:0000228': 63, 'Cellular Component::GO:0097525': 64, 'Cellular Component::GO:0098803': 65, 'Cellular Component::GO:0005774': 66, 'Cellular Component::GO:0005924': 67, 'Cellular Component::GO:0005740': 68, 'Cellular Component::GO:0000781': 69, 'Cellular Component::GO:0005769': 70, 'Cellular Component::GO:0031300': 71, 'Cellular Component::GO:0030133': 72, 'Cellular Component::GO:0032993': 73, 'Cellular Component::GO:0000793': 74, 'Cellular Component::GO:0031012': 75, 'Cellular Component::GO:0098796': 76, 'Cellular Component::GO:0005815': 77, 'Cellular Component::GO:0098794': 78, 'Cellular Component::GO:0005838': 79, 'Cellular Component::GO:0055037': 80, 'Cellular Component::GO:0043197': 81, 'Cellular Component::GO:0005834': 82, 'Cellular Component::GO:0070971': 83, 'Cellular Component::GO:0030426': 84, 'Cellular Component::GO:0005896': 85, 'Cellular Component::GO:0036064': 86, 'Cellular Component::GO:0043256': 87, 'Cellular Component::GO:0005604': 88, 'Cellular Component::GO:0005929': 89, 'Cellular Component::GO:0098636': 90, 'Cellular Component::GO:0030135': 91, 'Cellular Component::GO:0000152': 92, 'Cellular Component::GO:0030117': 93, 'Cellular Component::GO:0045334': 94, 'Cellular Component::GO:0030659': 95, 'Cellular Component::GO:0030424': 96, 'Cellular Component::GO:0000836': 97, 'Cellular Component::GO:0034451': 98, 'Cellular Component::GO:0005578': 99, 'Cellular Component::GO:0030141': 100, 'Cellular Component::GO:0019866': 101, 'Cellular Component::GO:0005746': 102, 'Cellular Component::GO:0005861': 103, 'Cellular Component::GO:0031464': 104, 'Cellular Component::GO:0009897': 105, 'Cellular Component::GO:0030897': 106, 'Cellular Component::GO:0070469': 107, 'Cellular Component::GO:0030136': 108, 'Cellular Component::GO:0030137': 109, 'Cellular Component::GO:0097452': 110, 'Cellular Component::GO:0009925': 111, 'Cellular Component::GO:0005759': 112, 'Cellular Component::GO:0098590': 113, 'Cellular Component::GO:0031970': 114, 'Cellular Component::GO:0099513': 115, 'Cellular Component::GO:0016942': 116, 'Cellular Component::GO:0005694': 117, 'Cellular Component::GO:0045271': 118, 'Cellular Component::GO:0005884': 119, 'Cellular Component::GO:0030176': 120, 'Cellular Component::GO:0005911': 121, 'Cellular Component::GO:0070161': 122, 'Cellular Component::GO:0005730': 123, 'Cellular Component::GO:0034993': 124, 'Cellular Component::GO:0005635': 125, 'Cellular Component::GO:0016328': 126, 'Cellular Component::GO:0005814': 127, 'Cellular Component::GO:0005813': 128, 'Cellular Component::GO:0012506': 129, 'Cellular Component::GO:0098797': 130, 'Cellular Component::GO:0005925': 131, 'Cellular Component::GO:0000785': 132, 'Cellular Component::GO:0097481': 133, 'Cellular Component::GO:0016234': 134, 'Cellular Component::GO:0031105': 135, 'Cellular Component::GO:0048770': 136, 'Cellular Component::GO:0034702': 137, 'Cellular Component::GO:0030130': 138, 'Cellular Component::GO:1902495': 139, 'Cellular Component::GO:0005840': 140, 'Cellular Component::GO:0005768': 141, 'Cellular Component::GO:0000780': 142, 'Cellular Component::GO:0045202': 143, 'Cellular Component::GO:0005681': 144, 'Cellular Component::GO:0036449': 145, 'Cellular Component::GO:0031968': 146, 'Cellular Component::GO:0044853': 147, 'Cellular Component::GO:0001669': 148, 'Cellular Component::GO:0015935': 149, 'Cellular Component::GO:0030139': 150, 'Cellular Component::GO:0000139': 151, 'Cellular Component::GO:1990234': 152, 'Cellular Component::GO:0000775': 153, 'Cellular Component::GO:0098798': 154, 'Cellular Component::GO:0030118': 155, 'Cellular Component::GO:0031225': 156, 'Cellular Component::GO:0000502': 157, 'Cellular Component::GO:0000323': 158, 'Cellular Component::GO:0005876': 159, 'Cellular Component::GO:0060076': 160, 'Cellular Component::GO:0071014': 161, 'Cellular Component::GO:0030173': 162, 'Cellular Component::GO:0005764': 163, 'Cellular Component::GO:0099572': 164, 'Cellular Component::GO:0097635': 165, 'Cellular Component::GO:0035145': 166, 'Cellular Component::GO:0034663': 167, 'Cellular Component::GO:0030425': 168, 'Cellular Component::GO:0035189': 169, 'Cellular Component::GO:0030662': 170, 'Cellular Component::GO:0031143': 171, 'Cellular Component::GO:0098852': 172, 'Cellular Component::GO:0005811': 173, 'Cellular Component::GO:0005891': 174, 'Cellular Component::GO:0034992': 175, 'Cellular Component::GO:0034686': 176, 'Cellular Component::GO:0000800': 177, 'Cellular Component::GO:0005875': 178, 'Cellular Component::GO:0016605': 179, 'Cellular Component::GO:0032391': 180, 'Cellular Component::GO:0034703': 181, 'Cellular Component::GO:0044304': 182, 'Cellular Component::GO:0000931': 183, 'Cellular Component::GO:0043202': 184, 'Cellular Component::GO:0005643': 185, 'Cellular Component::GO:0005901': 186, 'Cellular Component::GO:0034709': 187, 'Cellular Component::GO:0030688': 188, 'Cellular Component::GO:0005903': 189, 'Cellular Component::GO:0030529': 190, 'Cellular Component::GO:0030055': 191, 'Cellular Component::GO:0099501': 192, 'Cellular Component::GO:0099568': 193, 'Cellular Component::GO:0005819': 194, 'Cellular Component::GO:0000940': 195, 'Cellular Component::GO:0016607': 196, 'Cellular Component::GO:0010494': 197, 'Cellular Component::GO:0005798': 198, 'Cellular Component::GO:0005720': 199, 'Cellular Component::GO:0098862': 200, 'Cellular Component::GO:0019012': 201, 'Cellular Component::GO:1990851': 202, 'Cellular Component::GO:0043601': 203, 'Cellular Component::GO:0030686': 204, 'Cellular Component::GO:0031901': 205, 'Cellular Component::GO:0009898': 206, 'Cellular Component::GO:0005795': 207, 'Cellular Component::GO:0005839': 208, 'Cellular Component::GO:0016363': 209, 'Cellular Component::GO:0000786': 210, 'Cellular Component::GO:0016528': 211, 'Cellular Component::GO:0000922': 212, 'Cellular Component::GO:0030672': 213, 'Cellular Component::GO:0034098': 214, 'Cellular Component::GO:0005775': 215, 'Cellular Component::GO:0034774': 216, 'Cellular Component::GO:0016324': 217, 'Cellular Component::GO:0010008': 218, 'Cellular Component::GO:0033270': 219, 'Cellular Component::GO:0031984': 220, 'Cellular Component::GO:0031332': 221, 'Cellular Component::GO:0033503': 222, 'Cellular Component::GO:0045257': 223, 'Cellular Component::GO:0031234': 224, 'Cellular Component::GO:0034361': 225, 'Cellular Component::GO:0044309': 226, 'Cellular Component::GO:0034705': 227, 'Cellular Component::GO:0031902': 228, 'Cellular Component::GO:0097542': 229, 'Cellular Component::GO:0031082': 230, 'Cellular Component::GO:0097060': 231, 'Cellular Component::GO:0070187': 232, 'Cellular Component::GO:0000779': 233, 'Cellular Component::GO:1903293': 234, 'Cellular Component::GO:0098687': 235, 'Cellular Component::GO:0031463': 236, 'Cellular Component::GO:0071438': 237, 'Cellular Component::GO:0042272': 238, 'Cellular Component::GO:0060170': 239, 'Cellular Component::GO:0031903': 240, 'Cellular Component::GO:0000795': 241, 'Cellular Component::GO:0031301': 242, 'Cellular Component::GO:0043240': 243, 'Cellular Component::GO:0005765': 244, 'Cellular Component::GO:0072562': 245, 'Cellular Component::GO:0016323': 246, 'Cellular Component::GO:0099738': 247, 'Cellular Component::GO:0042734': 248, 'Cellular Component::GO:0042612': 249, 'Cellular Component::GO:0070461': 250, 'Cellular Component::GO:0005680': 251, 'Cellular Component::GO:0030008': 252, 'Cellular Component::GO:0033202': 253, 'Cellular Component::GO:0031461': 254, 'Cellular Component::GO:0031248': 255, 'Cellular Component::GO:0043205': 256, 'Cellular Component::GO:0005801': 257, 'Cellular Component::GO:0043679': 258, 'Cellular Component::GO:0031092': 259, 'Cellular Component::GO:0045211': 260, 'Cellular Component::GO:0005667': 261, 'Cellular Component::GO:0005802': 262, 'Cellular Component::GO:0017053': 263, 'Cellular Component::GO:0000407': 264, 'Cellular Component::GO:0030027': 265, 'Cellular Component::GO:0005741': 266, 'Cellular Component::GO:0008021': 267, 'Cellular Component::GO:0005916': 268, 'Cellular Component::GO:0036057': 269, 'Cellular Component::GO:0031362': 270, 'Cellular Component::GO:0032300': 271, 'Cellular Component::GO:0019897': 272, 'Cellular Component::GO:0044299': 273, 'Cellular Component::GO:0044300': 274, 'Cellular Component::GO:0097504': 275, 'Cellular Component::GO:0035327': 276, 'Cellular Component::GO:0000428': 277, 'Cellular Component::GO:0005881': 278, 'Cellular Component::GO:0072372': 279, 'Cellular Component::GO:0005905': 280, 'Cellular Component::GO:0044306': 281, 'Cellular Component::GO:0008023': 282, 'Cellular Component::GO:0036464': 283, 'Cellular Component::GO:0031228': 284, 'Cellular Component::GO:0015629': 285, 'Cellular Component::GO:0005930': 286, 'Cellular Component::GO:0001917': 287, 'Cellular Component::GO:0030285': 288, 'Cellular Component::GO:0036513': 289, 'Cellular Component::GO:0001741': 290, 'Cellular Component::GO:0042470': 291, 'Cellular Component::GO:0070603': 292, 'Cellular Component::GO:1902911': 293, 'Cellular Component::GO:0046658': 294, 'Cellular Component::GO:0043218': 295, 'Cellular Component::GO:0042622': 296, 'Cellular Component::GO:0022625': 297, 'Cellular Component::GO:0008280': 298, 'Cellular Component::GO:0032982': 299, 'Cellular Component::GO:0070382': 300, 'Cellular Component::GO:0030286': 301, 'Cellular Component::GO:0005796': 302, 'Cellular Component::GO:0032433': 303, 'Cellular Component::GO:0017101': 304, 'Cellular Component::GO:0098800': 305, 'Cellular Component::GO:0000788': 306, 'Cellular Component::GO:0030018': 307, 'Cellular Component::GO:0045179': 308, 'Cellular Component::GO:0043296': 309, 'Cellular Component::GO:0035770': 310, 'Cellular Component::GO:0097546': 311, 'Cellular Component::GO:0031965': 312, 'Cellular Component::GO:0036038': 313, 'Cellular Component::GO:0030880': 314, 'Cellular Component::GO:0022626': 315, 'Cellular Component::GO:0005915': 316, 'Cellular Component::GO:0030665': 317, 'Cellular Component::GO:0036452': 318, 'Cellular Component::GO:0000151': 319, 'Cellular Component::GO:0030119': 320, 'Cellular Component::GO:0005761': 321, 'Cellular Component::GO:0043203': 322, 'Cellular Component::GO:0000792': 323, 'Cellular Component::GO:0016529': 324, 'Cellular Component::GO:0030016': 325, 'Cellular Component::GO:0033116': 326, 'Cellular Component::GO:0032154': 327, 'Cellular Component::GO:0071565': 328, 'Cellular Component::GO:0031429': 329, 'Cellular Component::GO:0030427': 330, 'Cellular Component::GO:0097346': 331, 'Cellular Component::GO:0033011': 332, 'Cellular Component::GO:0030670': 333, 'Cellular Component::GO:0005771': 334, 'Cellular Component::GO:0031674': 335, 'Cellular Component::GO:0008076': 336, 'Cellular Component::GO:0031594': 337, 'Cellular Component::GO:0014069': 338, 'Cellular Component::GO:0032432': 339, 'Cellular Component::GO:0000783': 340, 'Cellular Component::GO:0061617': 341, 'Cellular Component::GO:0005883': 342, 'Cellular Component::GO:0043219': 343, 'Cellular Component::GO:0005892': 344, 'Cellular Component::GO:0005689': 345, 'Cellular Component::GO:0098573': 346, 'Cellular Component::GO:0042721': 347, 'Cellular Component::GO:0016235': 348, 'Cellular Component::GO:0043292': 349, 'Cellular Component::GO:0072686': 350, 'Cellular Component::GO:0008385': 351, 'Cellular Component::GO:0030663': 352, 'Cellular Component::GO:0031513': 353, 'Cellular Component::GO:0043034': 354, 'Cellular Component::GO:0097610': 355, 'Cellular Component::GO:0005777': 356, 'Cellular Component::GO:0016580': 357, 'Cellular Component::GO:0030532': 358, 'Cellular Component::GO:0043189': 359, 'Cellular Component::GO:0034708': 360, 'Cellular Component::GO:0031527': 361, 'Cellular Component::GO:0019867': 362, 'Cellular Component::GO:0032421': 363, 'Cellular Component::GO:0005762': 364, 'Cellular Component::GO:0031091': 365, 'Cellular Component::GO:0000791': 366, 'Cellular Component::GO:0005923': 367, 'Cellular Component::GO:0005736': 368, 'Cellular Component::GO:0000120': 369, 'Cellular Component::GO:0016591': 370, 'Cellular Component::GO:0000784': 371, 'Cellular Component::GO:0098858': 372, 'Cellular Component::GO:0031528': 373, 'Cellular Component::GO:0033017': 374, 'Cellular Component::GO:0070876': 375, 'Cellular Component::GO:0097014': 376, 'Cellular Component::GO:0000123': 377, 'Cellular Component::GO:0090575': 378, 'Cellular Component::GO:0033647': 379, 'Cellular Component::GO:0005577': 380, 'Cellular Component::GO:0044292': 381, 'Cellular Component::GO:0030673': 382, 'Cellular Component::GO:0045259': 383, 'Cellular Component::GO:0019773': 384, 'Cellular Component::GO:0030666': 385, 'Cellular Component::GO:0031983': 386, 'Cellular Component::GO:0031256': 387, 'Cellular Component::GO:0032580': 388, 'Cellular Component::GO:0005788': 389, 'Cellular Component::GO:0000145': 390, 'Cellular Component::GO:0030496': 391, 'Cellular Component::GO:0000308': 392, 'Cellular Component::GO:0033177': 393, 'Cellular Component::GO:0045178': 394, 'Cellular Component::GO:0031515': 395, 'Cellular Component::GO:0043596': 396, 'Cellular Component::GO:0030658': 397, 'Cellular Component::GO:0098562': 398, 'Cellular Component::GO:0048475': 399, 'Cellular Component::GO:0030863': 400, 'Cellular Component::GO:1990391': 401, 'Cellular Component::GO:0097038': 402, 'Cellular Component::GO:0016327': 403, 'Cellular Component::GO:0042383': 404, 'Cellular Component::GO:0000777': 405, 'Cellular Component::GO:0014704': 406, 'Cellular Component::GO:0031011': 407, 'Cellular Component::GO:0015934': 408, 'Cellular Component::GO:0044609': 409, 'Cellular Component::GO:0097165': 410, 'Cellular Component::GO:0000794': 411, 'Cellular Component::GO:0042382': 412, 'Cellular Component::GO:0031501': 413, 'Cellular Component::GO:0070160': 414, 'Cellular Component::GO:0005657': 415, 'Cellular Component::GO:0005581': 416, 'Cellular Component::GO:0030660': 417, 'Cellular Component::GO:0016592': 418, 'Cellular Component::GO:0005793': 419, 'Cellular Component::GO:0000315': 420, 'Cellular Component::GO:0032588': 421, 'Cellular Component::GO:0030877': 422, 'Cellular Component::GO:0005902': 423, 'Cellular Component::GO:0000314': 424, 'Cellular Component::GO:0005791': 425, 'Cellular Component::GO:0098552': 426, 'Cellular Component::GO:0031258': 427, 'Cellular Component::GO:0030915': 428, 'Cellular Component::GO:0044298': 429, 'Cellular Component::GO:0043209': 430, 'Cellular Component::GO:0015030': 431, 'Cellular Component::GO:0032279': 432, 'Cellular Component::GO:0034045': 433, 'Cellular Component::GO:0043195': 434, 'Cellular Component::GO:0016461': 435, 'Cellular Component::GO:0070822': 436, 'Cellular Component::GO:0000307': 437, 'Cellular Component::GO:0005851': 438, 'Cellular Component::GO:0002102': 439, 'Cellular Component::GO:0045335': 440, 'Cellular Component::GO:0016011': 441, 'Cellular Component::GO:0030057': 442, 'Cellular Component::GO:0042641': 443, 'Cellular Component::GO:0031985': 444, 'Cellular Component::GO:0030175': 445, 'Cellular Component::GO:0044665': 446, 'Cellular Component::GO:0031502': 447, 'Cellular Component::GO:0016459': 448, 'Cellular Component::GO:0001725': 449, 'Cellular Component::GO:0044391': 450, 'Cellular Component::GO:0031514': 451, 'Cellular Component::GO:0090665': 452, 'Cellular Component::GO:0042645': 453, 'Cellular Component::GO:0016471': 454, 'Cellular Component::GO:0000118': 455, 'Cellular Component::GO:0031233': 456, 'Cellular Component::GO:0030864': 457, 'Cellular Component::GO:0002080': 458, 'Cellular Component::GO:0016281': 459, 'Cellular Component::GO:0032153': 460, 'Cellular Component::GO:0034706': 461, 'Cellular Component::GO:0034366': 462, 'Cellular Component::GO:0032281': 463, 'Cellular Component::GO:0071682': 464, 'Cellular Component::GO:0035861': 465, 'Cellular Component::GO:0098576': 466, 'Cellular Component::GO:0030140': 467, 'Cellular Component::GO:0032994': 468, 'Cellular Component::GO:0042613': 469, 'Cellular Component::GO:0032389': 470, 'Cellular Component::GO:0031616': 471, 'Cellular Component::GO:0030017': 472, 'Cellular Component::GO:0000313': 473, 'Cellular Component::GO:0044798': 474, 'Cellular Component::GO:0005852': 475, 'Cellular Component::GO:0000778': 476, 'Cellular Component::GO:0071141': 477, 'Cellular Component::GO:0000125': 478, 'Cellular Component::GO:0030312': 479, 'Cellular Component::GO:0005782': 480, 'Cellular Component::GO:0005719': 481, 'Cellular Component::GO:0008282': 482, 'Cellular Component::GO:0005686': 483, 'Cellular Component::GO:0034704': 484, 'Cellular Component::GO:0032809': 485, 'Cellular Component::GO:0005637': 486, 'Cellular Component::GO:0033178': 487, 'Cellular Component::GO:0071013': 488, 'Cellular Component::GO:0005669': 489, 'Cellular Component::GO:0071203': 490, 'Cellular Component::GO:0070938': 491, 'Cellular Component::GO:0098644': 492, 'Cellular Component::GO:0005847': 493, 'Cellular Component::GO:0031262': 494, 'Cellular Component::GO:0008274': 495, 'Cellular Component::GO:0005882': 496, 'Cellular Component::GO:0030684': 497, 'Cellular Component::GO:0031941': 498, 'Cellular Component::GO:0044217': 499, 'Cellular Component::GO:0043198': 500, 'Cellular Component::GO:0045171': 501, 'Cellular Component::GO:0005790': 502, 'Cellular Component::GO:0035097': 503, 'Cellular Component::GO:0046930': 504, 'Cellular Component::GO:0005641': 505, 'Cellular Component::GO:0001740': 506, 'Cellular Component::GO:0072589': 507, 'Cellular Component::GO:0097517': 508, 'Cellular Component::GO:0097632': 509, 'Cellular Component::GO:0005747': 510, 'Cellular Component::GO:0005684': 511, 'Cellular Component::GO:0071782': 512, 'Cellular Component::GO:0031519': 513, 'Cellular Component::GO:0030677': 514, 'Cellular Component::GO:0031312': 515, 'Cellular Component::GO:0070820': 516, 'Cellular Component::GO:0045275': 517, 'Cellular Component::GO:0033093': 518, 'Cellular Component::GO:0044327': 519, 'Cellular Component::GO:0005776': 520, 'Cellular Component::GO:0098802': 521, 'Cellular Component::GO:0034719': 522, 'Cellular Component::GO:0031430': 523, 'Cellular Component::GO:0031093': 524, 'Cellular Component::GO:0032589': 525, 'Cellular Component::GO:0032002': 526, 'Cellular Component::GO:0016600': 527, 'Cellular Component::GO:0036126': 528, 'Cellular Component::GO:0032587': 529, 'Cellular Component::GO:0080008': 530, 'Cellular Component::GO:0031094': 531, 'Cellular Component::GO:0008328': 532, 'Cellular Component::GO:0033391': 533, 'Cellular Component::GO:0031201': 534, 'Cellular Component::GO:0045120': 535, 'Cellular Component::GO:0071339': 536, 'Cellular Component::GO:1990761': 537, 'Cellular Component::GO:0033179': 538, 'Cellular Component::GO:0032420': 539, 'Cellular Component::GO:0060205': 540, 'Cellular Component::GO:0008305': 541, 'Cellular Component::GO:1904949': 542, 'Cellular Component::GO:0045239': 543, 'Cellular Component::GO:0042579': 544, 'Cellular Component::GO:0005927': 545, 'Cellular Component::GO:0002139': 546, 'Cellular Component::GO:0042611': 547, 'Cellular Component::GO:0032009': 548, 'Cellular Component::GO:0044611': 549, 'Cellular Component::GO:0005778': 550, 'Cellular Component::GO:0016514': 551, 'Cellular Component::GO:1902493': 552, 'Cellular Component::GO:0030478': 553, 'Cellular Component::GO:0005749': 554, 'Cellular Component::GO:0038201': 555, 'Cellular Component::GO:0097651': 556, 'Cellular Component::GO:0001520': 557, 'Cellular Component::GO:0005797': 558, 'Cellular Component::GO:0000803': 559, 'Cellular Component::GO:0005639': 560, 'Cellular Component::GO:0005682': 561, 'Cellular Component::GO:0005947': 562, 'Cellular Component::GO:0031526': 563, 'Cellular Component::GO:0098799': 564, 'Cellular Component::GO:0005675': 565, 'Cellular Component::GO:0045281': 566, 'Cellular Component::GO:0097136': 567, 'Cellular Component::GO:0048188': 568, 'Cellular Component::GO:0030894': 569, 'Cellular Component::GO:0042405': 570, 'Cellular Component::GO:0070419': 571, 'Cellular Component::GO:0044215': 572, 'Cellular Component::GO:0046581': 573, 'Cellular Component::GO:0005863': 574, 'Cellular Component::GO:0030990': 575, 'Cellular Component::GO:0030891': 576, 'Cellular Component::GO:0031307': 577, 'Cellular Component::GO:0033180': 578, 'Cellular Component::GO:0034362': 579, 'Cellular Component::GO:0031231': 580, 'Cellular Component::GO:0061702': 581, 'Cellular Component::GO:0010369': 582, 'Cellular Component::GO:0000815': 583, 'Cellular Component::GO:0034358': 584, 'Cellular Component::GO:1990667': 585, 'Cellular Component::GO:0030125': 586, 'Cellular Component::GO:1990777': 587, 'Cellular Component::GO:0070069': 588, 'Cellular Component::GO:0036379': 589, 'Cellular Component::GO:0005671': 590, 'Cellular Component::GO:0044307': 591, 'Cellular Component::GO:0005732': 592, 'Cellular Component::GO:0005606': 593, 'Cellular Component::GO:0055038': 594, 'Cellular Component::GO:0000932': 595, 'Cellular Component::GO:0034457': 596, 'Cellular Component::GO:0005724': 597, 'Cellular Component::GO:0000243': 598, 'Cellular Component::GO:0005584': 599, 'Cellular Component::GO:0032591': 600, 'Cellular Component::GO:0016035': 601, 'Cellular Component::GO:0030681': 602, 'Cellular Component::GO:0034464': 603, 'Cellular Component::GO:0043194': 604, 'Cellular Component::GO:0032839': 605, 'Cellular Component::GO:0045263': 606, 'Cellular Component::GO:0097539': 607, 'Cellular Component::GO:0098651': 608, 'Cellular Component::GO:0045009': 609, 'Cellular Component::GO:0022624': 610, 'Cellular Component::GO:1990454': 611, 'Cellular Component::GO:0034707': 612, 'Cellular Component::GO:0035748': 613, 'Cellular Component::GO:0005865': 614, 'Cellular Component::GO:0001891': 615, 'Cellular Component::GO:0036501': 616, 'Cellular Component::GO:0032590': 617, 'Cellular Component::GO:0001652': 618, 'Cellular Component::GO:0016469': 619, 'Cellular Component::GO:0070557': 620, 'Cellular Component::GO:0033018': 621, 'Cellular Component::GO:0035631': 622, 'Cellular Component::GO:1990005': 623, 'Cellular Component::GO:0043511': 624, 'Cellular Component::GO:0010009': 625, 'Cellular Component::GO:0035267': 626, 'Cellular Component::GO:0035859': 627, 'Cellular Component::GO:0071020': 628, 'Cellular Component::GO:0033268': 629, 'Cellular Component::GO:0030964': 630, 'Cellular Component::GO:1990879': 631, 'Cellular Component::GO:1990696': 632, 'Cellular Component::GO:0043656': 633, 'Cellular Component::GO:0070083': 634, 'Cellular Component::GO:0071541': 635, 'Cellular Component::GO:0005832': 636, 'Cellular Component::GO:0031232': 637, 'Cellular Component::GO:0032998': 638, 'Cellular Component::GO:0071546': 639, 'Cellular Component::GO:0032806': 640, 'Cellular Component::GO:0000346': 641, 'Cellular Component::GO:0005640': 642, 'Cellular Component::GO:0090576': 643, 'Cellular Component::GO:0033263': 644, 'Cellular Component::GO:0005665': 645, 'Cellular Component::GO:0005672': 646, 'Cellular Component::GO:0005753': 647, 'Cellular Component::GO:0000110': 648, 'Cellular Component::GO:0097225': 649, 'Cellular Component::GO:0042555': 650, 'Cellular Component::GO:0034456': 651, 'Cellular Component::GO:0034385': 652, 'Cellular Component::GO:0000124': 653, 'Cellular Component::GO:0061700': 654, 'Cellular Component::GO:1990604': 655, 'Cellular Component::GO:0071011': 656, 'Cellular Component::GO:0051233': 657, 'Cellular Component::GO:0002116': 658, 'Cellular Component::GO:1990909': 659, 'Cellular Component::GO:0097025': 660, 'Cellular Component::GO:0042599': 661, 'Cellular Component::GO:0031095': 662, 'Cellular Component::GO:0000813': 663, 'Cellular Component::GO:0019005': 664, 'Cellular Component::GO:0017059': 665, 'Cellular Component::GO:0071439': 666, 'Cellular Component::GO:0043259': 667, 'Cellular Component::GO:0000127': 668, 'Cellular Component::GO:0030992': 669, 'Cellular Component::GO:0060201': 670, 'Cellular Component::GO:0022627': 671, 'Cellular Component::GO:0005844': 672, 'Cellular Component::GO:0030687': 673, 'Cellular Component::GO:0097449': 674, 'Cellular Component::GO:0097342': 675, 'Cellular Component::GO:0019907': 676, 'Cellular Component::GO:0031304': 677, 'Cellular Component::GO:0005965': 678, 'Cellular Component::GO:0005942': 679, 'Cellular Component::GO:0035102': 680, 'Cellular Component::GO:0046540': 681, 'Cellular Component::GO:0048500': 682, 'Cellular Component::GO:0016460': 683, 'Cellular Component::GO:0045254': 684, 'Cellular Component::GO:1902560': 685, 'Cellular Component::GO:0042581': 686, 'Cellular Component::GO:1902554': 687, 'Cellular Component::GO:0070688': 688, 'Cellular Component::GO:0005828': 689, 'Cellular Component::GO:0031298': 690, 'Cellular Component::GO:0098553': 691, 'Cellular Component::GO:0030056': 692, 'Cellular Component::GO:0042582': 693, 'Cellular Component::GO:0032584': 694, 'Cellular Component::GO:0032592': 695, 'Cellular Component::GO:0060091': 696, 'Cellular Component::GO:0031305': 697, 'Cellular Component::GO:0035355': 698, 'Cellular Component::GO:0071010': 699, 'Cellular Component::GO:0090543': 700, 'Cellular Component::GO:0072669': 701, 'Cellular Component::GO:0031672': 702, 'Cellular Component::GO:0031265': 703, 'Cellular Component::GO:0012510': 704, 'Cellular Component::GO:0032797': 705, 'Cellular Component::GO:0005588': 706, 'Cellular Component::GO:0016282': 707, 'Cellular Component::GO:0042719': 708, 'Cellular Component::GO:0031045': 709, 'Cellular Component::GO:0048786': 710, 'Cellular Component::GO:0031907': 711, 'Cellular Component::GO:0042629': 712, 'Cellular Component::GO:0000214': 713, 'Cellular Component::GO:0071797': 714, 'Cellular Component::GO:0008278': 715, 'Cellular Component::GO:0031083': 716, 'Cellular Component::GO:0030313': 717, 'Cellular Component::GO:0098827': 718, 'Cellular Component::GO:0030914': 719, 'Cellular Component::GO:0030123': 720, 'Cellular Component::GO:0097526': 721, 'Cellular Component::GO:0005779': 722, 'Cellular Component::GO:1990726': 723, 'Cellular Component::GO:0005697': 724, 'Cellular Component::GO:0090533': 725, 'Cellular Component::GO:0097381': 726, 'Cellular Component::GO:0097228': 727, 'Cellular Component::GO:0031588': 728, 'Cellular Component::GO:0070449': 729, 'Cellular Component::GO:0035253': 730, 'Cellular Component::GO:0017119': 731, 'Cellular Component::GO:0008290': 732, 'Cellular Component::GO:0042627': 733, 'Cellular Component::GO:0030134': 734, 'Cellular Component::GO:0000802': 735, 'Cellular Component::GO:0070847': 736, 'Cellular Component::GO:0097233': 737, 'Cellular Component::GO:0034991': 738, 'Cellular Component::GO:0005943': 739, 'Cellular Component::GO:0044233': 740, 'Cellular Component::GO:0014731': 741, 'Cellular Component::GO:0005767': 742, 'Cellular Component::GO:0014701': 743, 'Cellular Component::GO:0060198': 744, 'Cellular Component::GO:0012507': 745, 'Cellular Component::GO:0005685': 746, 'Cellular Component::GO:0090545': 747, 'Cellular Component::GO:0008091': 748, 'Cellular Component::GO:0034518': 749, 'Cellular Component::GO:0036020': 750, 'Cellular Component::GO:0016010': 751, 'Cellular Component::GO:0000177': 752, 'Cellular Component::GO:0000306': 753, 'Cellular Component::GO:0005763': 754, 'Cellular Component::GO:1902711': 755, 'Cellular Component::GO:0090544': 756, 'Cellular Component::GO:0070552': 757, 'Cellular Component::GO:0071564': 758, 'Cellular Component::GO:0008024': 759, 'Cellular Component::GO:0018995': 760, 'Cellular Component::GO:0005953': 761, 'Cellular Component::GO:0097541': 762, 'Cellular Component::GO:0009295': 763, 'Cellular Component::GO:0000235': 764, 'Cellular Component::GO:0035578': 765, 'Cellular Component::GO:0044224': 766, 'Cellular Component::GO:0031414': 767, 'Cellular Component::GO:0000439': 768, 'Cellular Component::GO:0042101': 769, 'Cellular Component::GO:0097232': 770, 'Cellular Component::GO:0030669': 771, 'Cellular Component::GO:0045095': 772, 'Cellular Component::GO:0031597': 773, 'Cellular Component::GO:0060293': 774, 'Cellular Component::GO:0071664': 775, 'Cellular Component::GO:0033010': 776, 'Cellular Component::GO:0000811': 777, 'Cellular Component::GO:0042585': 778, 'Cellular Component::GO:0032541': 779, 'Cellular Component::GO:0071598': 780, 'Cellular Component::GO:0030126': 781, 'Cellular Component::GO:0032585': 782, 'Cellular Component::GO:0072559': 783, 'Cellular Component::GO:0032059': 784, 'Cellular Component::GO:0043186': 785, 'Cellular Component::GO:0090568': 786, 'Cellular Component::GO:1990752': 787, 'Cellular Component::GO:0034678': 788, 'Cellular Component::GO:0036157': 789, 'Cellular Component::GO:0033150': 790, 'Cellular Component::GO:0070288': 791, 'Cellular Component::GO:0044194': 792, 'Cellular Component::GO:0071004': 793, 'Cellular Component::GO:0071437': 794, 'Cellular Component::GO:1902710': 795, 'Cellular Component::GO:1902562': 796, 'Cellular Component::GO:0000178': 797, 'Cellular Component::GO:0005845': 798, 'Cellular Component::GO:1903349': 799, 'Cellular Component::GO:0005664': 800, 'Cellular Component::GO:0000242': 801, 'Cellular Component::GO:0033565': 802, 'Cellular Component::GO:0032838': 803, 'Cellular Component::GO:0005687': 804, 'Cellular Component::GO:0043083': 805, 'Cellular Component::GO:0071008': 806, 'Cellular Component::GO:0042824': 807, 'Cellular Component::GO:0030314': 808, 'Cellular Component::GO:0031466': 809, 'Cellular Component::GO:0032280': 810, 'Cellular Component::GO:0016272': 811, 'Cellular Component::GO:0030870': 812, 'Cellular Component::GO:0005655': 813, 'Cellular Component::GO:0071953': 814, 'Cellular Component::GO:0070993': 815, 'Cellular Component::GO:1990023': 816, 'Cellular Component::GO:0017146': 817, 'Cellular Component::GO:0033176': 818, 'Cellular Component::GO:0005589': 819, 'Cellular Component::GO:0097197': 820, 'Cellular Component::GO:0031933': 821, 'Cellular Component::GO:0005721': 822, 'Cellular Component::GO:0032426': 823, 'Cellular Component::GO:1903095': 824, 'Cellular Component::GO:0005726': 825, 'Cellular Component::GO:0030893': 826, 'Cellular Component::GO:0005614': 827, 'Cellular Component::GO:0042825': 828, 'Cellular Component::GO:0098533': 829, 'Cellular Component::GO:0032040': 830, 'Cellular Component::GO:0035749': 831, 'Cellular Component::GO:0097433': 832, 'Cellular Component::GO:0097208': 833, 'Cellular Component::GO:0005833': 834, 'Cellular Component::GO:0005587': 835, 'Cellular Component::GO:0035061': 836, 'Cellular Component::GO:0043657': 837, 'Cellular Component::GO:0034683': 838, 'Cellular Component::GO:0098642': 839, 'Cellular Component::GO:0032302': 840, 'Cellular Component::GO:0033276': 841, 'Cellular Component::GO:1990597': 842, 'Cellular Component::GO:0070545': 843, 'Cellular Component::GO:0070939': 844, 'Cellular Component::GO:1902937': 845, 'Cellular Component::GO:0035371': 846, 'Cellular Component::GO:0005868': 847, 'Cellular Component::GO:0008541': 848, 'Cellular Component::GO:0005952': 849, 'Cellular Component::GO:0030015': 850, 'Cellular Component::GO:0000506': 851, 'Cellular Component::GO:0042589': 852, 'Cellular Component::GO:0030934': 853, 'Cellular Component::GO:0002133': 854, 'Cellular Component::GO:0000445': 855, 'Cellular Component::GO:1903767': 856, 'Cellular Component::GO:0000322': 857, 'Cellular Component::GO:0097209': 858, 'Cellular Component::GO:0000347': 859, 'Cellular Component::GO:0031251': 860, 'Cellular Component::GO:0035363': 861, 'Cellular Component::GO:0000805': 862, 'Cellular Component::GO:0005666': 863, 'Cellular Component::GO:0005766': 864, 'Cellular Component::GO:0019814': 865, 'Cellular Component::GO:0044354': 866, 'Cellular Component::GO:0097057': 867, 'Cellular Component::GO:0005744': 868, 'Cellular Component::GO:0001772': 869, 'Cellular Component::GO:0005742': 870, 'Cellular Component::GO:0005662': 871, 'Cellular Component::GO:0005827': 872, 'Cellular Component::GO:0031512': 873, 'Cellular Component::GO:0031314': 874, 'Cellular Component::GO:0042587': 875, 'Cellular Component::GO:0001673': 876, 'Cellular Component::GO:0030131': 877, 'Cellular Component::GO:0043220': 878, 'Cellular Component::GO:1990077': 879, 'Cellular Component::GO:0072357': 880, 'Cellular Component::GO:0001739': 881, 'Cellular Component::GO:0002199': 882, 'Cellular Component::GO:0033186': 883, 'Cellular Component::GO:0000109': 884, 'Cellular Component::GO:0072563': 885, 'Cellular Component::GO:0002141': 886, 'Cellular Component::GO:0000923': 887, 'Cellular Component::GO:0031595': 888, 'Cellular Component::GO:0043190': 889, 'Cellular Component::GO:0033655': 890, 'Cellular Component::GO:0016442': 891, 'Cellular Component::GO:0044295': 892, 'Cellular Component::GO:0036195': 893, 'Cellular Component::GO:0035869': 894, 'Cellular Component::GO:0032311': 895, 'Cellular Component::GO:0016581': 896, 'Cellular Component::GO:0043512': 897, 'Cellular Component::GO:0036396': 898, 'Cellular Component::GO:0000164': 899, 'Cellular Component::GO:0070761': 900, 'Cellular Component::GO:0030121': 901, 'Cellular Component::GO:0016012': 902, 'Cellular Component::GO:0032156': 903, 'Cellular Component::GO:0001674': 904, 'Cellular Component::GO:0000153': 905, 'Cellular Component::GO:0071986': 906, 'Cellular Component::GO:0072546': 907, 'Cellular Component::GO:0072558': 908, 'Cellular Component::GO:0030120': 909, 'Cellular Component::GO:0034684': 910, 'Cellular Component::GO:0071547': 911, 'Cellular Component::GO:0031906': 912, 'Cellular Component::GO:0030991': 913, 'Cellular Component::GO:0042406': 914, 'Cellular Component::GO:0070578': 915, 'Cellular Component::GO:1990907': 916, 'Cellular Component::GO:0031235': 917, 'Cellular Component::GO:0033290': 918, 'Cellular Component::GO:0008540': 919, 'Cellular Component::GO:0043541': 920, 'Cellular Component::GO:0071065': 921, 'Cellular Component::GO:0031088': 922, 'Cellular Component::GO:0005922': 923, 'Cellular Component::GO:0061200': 924, 'Cellular Component::GO:0000814': 925, 'Cellular Component::GO:0005859': 926, 'Cellular Component::GO:0005850': 927, 'Cellular Component::GO:0070033': 928, 'Cellular Component::GO:0030062': 929, 'Cellular Component::GO:0000782': 930, 'Cellular Component::GO:0008074': 931, 'Cellular Component::GO:0031209': 932, 'Cellular Component::GO:0030128': 933, 'Cellular Component::GO:0098575': 934, 'Cellular Component::GO:0035101': 935, 'Cellular Component::GO:0045273': 936, 'Cellular Component::GO:0005968': 937, 'Cellular Component::GO:0005751': 938, 'Cellular Component::GO:0001527': 939, 'Cellular Component::GO:0090537': 940, 'Cellular Component::GO:0097441': 941, 'Cellular Component::GO:0031932': 942, 'Cellular Component::GO:0044326': 943, 'Cellular Component::GO:0000015': 944, 'Cellular Component::GO:0097059': 945, 'Cellular Component::GO:0042575': 946, 'Cellular Component::GO:0005785': 947, 'Cellular Component::GO:1990635': 948, 'Cellular Component::GO:0030904': 949, 'Cellular Component::GO:0000421': 950, 'Cellular Component::GO:0042720': 951, 'Cellular Component::GO:0048269': 952, 'Cellular Component::GO:0034364': 953, 'Cellular Component::GO:0005921': 954, 'Cellular Component::GO:0034515': 955, 'Cellular Component::GO:0034687': 956, 'Cellular Component::GO:0000812': 957, 'Cellular Component::GO:1990357': 958, 'Cellular Component::GO:0060053': 959, 'Cellular Component::GO:0043625': 960, 'Cellular Component::GO:0000176': 961, 'Cellular Component::GO:0033185': 962, 'Cellular Component::GO:0001518': 963, 'Cellular Component::GO:0032021': 964, 'Cellular Component::GO:0044232': 965, 'Cellular Component::GO:0002189': 966, 'Cellular Component::GO:0070531': 967, 'Cellular Component::GO:0031465': 968, 'Cellular Component::GO:0005605': 969, 'Cellular Component::GO:0070652': 970, 'Cellular Component::GO:0071819': 971, 'Cellular Component::GO:0045495': 972, 'Cellular Component::GO:0045283': 973, 'Cellular Component::GO:0005583': 974, 'Cellular Component::GO:1902555': 975, 'Cellular Component::GO:0060199': 976, 'Cellular Component::GO:0044216': 977, 'Cellular Component::GO:0005964': 978, 'Cellular Component::GO:0097450': 979, 'Cellular Component::GO:0005688': 980, 'Cellular Component::GO:0034666': 981, 'Cellular Component::GO:0097648': 982, 'Cellular Component::GO:0030132': 983, 'Cellular Component::GO:0016593': 984, 'Cellular Component::GO:0034679': 985, 'Cellular Component::GO:0034715': 986, 'Cellular Component::GO:1990037': 987, 'Cellular Component::GO:1990393': 988, 'Cellular Component::GO:0005610': 989, 'Cellular Component::GO:0031372': 990, 'Cellular Component::GO:0042567': 991, 'Cellular Component::GO:0005869': 992, 'Cellular Component::GO:0045277': 993, 'Cellular Component::GO:0033256': 994, 'Cellular Component::GO:0005677': 995, 'Cellular Component::GO:0005592': 996, 'Cellular Component::GO:0030112': 997, 'Cellular Component::GO:0005955': 998, 'Cellular Component::GO:0045252': 999, 'Cellular Component::GO:1990712': 1000, 'Cellular Component::GO:0005760': 1001, 'Cellular Component::GO:0019908': 1002, 'Cellular Component::GO:0045025': 1003, 'Cellular Component::GO:0032545': 1004, 'Cellular Component::GO:0036457': 1005, 'Cellular Component::GO:0031428': 1006, 'Cellular Component::GO:0098536': 1007, 'Cellular Component::GO:0008250': 1008, 'Cellular Component::GO:0071942': 1009, 'Cellular Component::GO:0005858': 1010, 'Cellular Component::GO:0005786': 1011, 'Cellular Component::GO:0071001': 1012, 'Cellular Component::GO:1990589': 1013, 'Cellular Component::GO:0005945': 1014, 'Cellular Component::GO:0031904': 1015, 'Cellular Component::GO:0044614': 1016, 'Cellular Component::GO:0036128': 1017, 'Cellular Component::GO:0034066': 1018, 'Cellular Component::GO:0042588': 1019, 'Cellular Component::GO:0072687': 1020, 'Cellular Component::GO:0000938': 1021, 'Cellular Component::GO:0045261': 1022, 'Cellular Component::GO:0031230': 1023, 'Cellular Component::GO:0071821': 1024, 'Cellular Component::GO:0051286': 1025, 'Cellular Component::GO:0043527': 1026, 'Cellular Component::GO:0000806': 1027, 'Cellular Component::GO:0031416': 1028, 'Cellular Component::GO:0043196': 1029, 'Cellular Component::GO:0008622': 1030, 'Cellular Component::GO:0098643': 1031, 'Cellular Component::GO:0032777': 1032, 'Cellular Component::GO:0097418': 1033, 'Cellular Component::GO:0070044': 1034, 'Cellular Component::GO:0031417': 1035, 'Cellular Component::GO:0034363': 1036, 'Cellular Component::GO:0036019': 1037, 'Cellular Component::GO:0016589': 1038, 'Cellular Component::GO:0000796': 1039, 'Cellular Component::GO:0031264': 1040, 'Cellular Component::GO:0097629': 1041, 'Cellular Component::GO:0045240': 1042, 'Cellular Component::GO:1990131': 1043, 'Cellular Component::GO:0070776': 1044, 'Cellular Component::GO:0030127': 1045, 'Cellular Component::GO:0005889': 1046, 'Cellular Component::GO:0060077': 1047, 'Cellular Component::GO:0005885': 1048, 'Cellular Component::GO:0008623': 1049, 'Cellular Component::GO:0005642': 1050, 'Cellular Component::GO:0005862': 1051, 'Cellular Component::GO:0031371': 1052, 'Cellular Component::GO:0090498': 1053, 'Cellular Component::GO:0072379': 1054, 'Cellular Component::GO:0005579': 1055, 'Cellular Component::GO:0000798': 1056, 'Cellular Component::GO:0030868': 1057, 'Cellular Component::GO:0055087': 1058, 'Cellular Component::GO:0001939': 1059, 'Cellular Component::GO:0071062': 1060, 'Cellular Component::GO:0097513': 1061, 'Cellular Component::GO:0035354': 1062, 'Cellular Component::GO:0031467': 1063, 'Cellular Component::GO:0035098': 1064, 'Cellular Component::GO:0044352': 1065, 'Cellular Component::GO:0016602': 1066, 'Cellular Component::GO:0032045': 1067, 'Cellular Component::GO:0031010': 1068, 'Cellular Component::GO:0045298': 1069, 'Cellular Component::GO:0005658': 1070, 'Cellular Component::GO:0019815': 1071, 'Cellular Component::GO:0097386': 1072, 'Cellular Component::GO:0030430': 1073, 'Cellular Component::GO:0005818': 1074, 'Cellular Component::GO:0005602': 1075, 'Cellular Component::GO:0000138': 1076, 'Cellular Component::GO:0070032': 1077, 'Cellular Component::GO:0071256': 1078, 'Cellular Component::GO:0070937': 1079, 'Cellular Component::GO:0000799': 1080, 'Cellular Component::GO:0098647': 1081, 'Cellular Component::GO:1990111': 1082, 'Cellular Component::GO:0070852': 1083, 'Cellular Component::GO:0000808': 1084, 'Cellular Component::GO:0002945': 1085, 'Cellular Component::GO:0032449': 1086, 'Cellular Component::GO:0000835': 1087, 'Cellular Component::GO:0098554': 1088, 'Cellular Component::GO:0035838': 1089, 'Cellular Component::GO:0098645': 1090, 'Cellular Component::GO:0097454': 1091, 'Cellular Component::GO:0097440': 1092, 'Cellular Component::GO:0005846': 1093, 'Cellular Component::GO:0033162': 1094, 'Cellular Component::GO:0097443': 1095, 'Cellular Component::GO:0061202': 1096, 'Cellular Component::GO:0097413': 1097, 'Cellular Component::GO:0070695': 1098, 'Cellular Component::GO:0031931': 1099, 'Cellular Component::GO:0008043': 1100, 'Cellular Component::GO:0097470': 1101, 'Cellular Component::GO:0005873': 1102, 'Cellular Component::GO:0071817': 1103, 'Cellular Component::GO:0031085': 1104, 'Cellular Component::GO:0033646': 1105, 'Cellular Component::GO:0072487': 1106, 'Cellular Component::GO:0033257': 1107, 'Cellular Component::GO:0005750': 1108, 'Cellular Component::GO:0036056': 1109, 'Cellular Component::GO:0000275': 1110, 'Cellular Component::GO:0035692': 1111, 'Cellular Component::GO:0005900': 1112, 'Cellular Component::GO:0033063': 1113, 'Cellular Component::GO:0036398': 1114, 'Cellular Component::GO:0097058': 1115, 'Cellular Component::GO:0000939': 1116, 'Cellular Component::GO:0042583': 1117, 'Cellular Component::GO:0046696': 1118, 'Cellular Component::GO:0070436': 1119, 'Cellular Component::GO:0005784': 1120, 'Cellular Component::GO:0032593': 1121, 'Cellular Component::GO:0035032': 1122, 'Cellular Component::GO:0097227': 1123, 'Cellular Component::GO:0036338': 1124, 'Cellular Component::GO:0030867': 1125, 'Cellular Component::GO:0070369': 1126, 'Cellular Component::GO:0001931': 1127, 'Cellular Component::GO:0042571': 1128, 'Cellular Component::GO:0071458': 1129, 'Cellular Component::GO:0031229': 1130, 'Cellular Component::GO:0032983': 1131, 'Cellular Component::GO:0000220': 1132, 'Cellular Component::GO:0098559': 1133, 'Cellular Component::GO:0042765': 1134, 'Cellular Component::GO:1990316': 1135, 'Cellular Component::GO:1990723': 1136, 'Cellular Component::GO:0030981': 1137, 'Cellular Component::GO:0032116': 1138, 'Cellular Component::GO:0097486': 1139, 'Cellular Component::GO:0070554': 1140, 'Cellular Component::GO:0045180': 1141, 'Cellular Component::GO:0044530': 1142, 'Cellular Component::GO:0036454': 1143, 'Cellular Component::GO:0005879': 1144, 'Cellular Component::GO:0031254': 1145, 'Cellular Component::GO:0031084': 1146, 'Cellular Component::GO:0099569': 1147, 'Cellular Component::GO:0072588': 1148, 'Cellular Component::GO:0097451': 1149, 'Cellular Component::GO:0034271': 1150, 'Cellular Component::GO:0020018': 1151, 'Cellular Component::GO:0005971': 1152, 'Cellular Component::GO:0035985': 1153, 'Cellular Component::GO:0034677': 1154, 'Cellular Component::GO:0098592': 1155, 'Cellular Component::GO:0005890': 1156, 'Cellular Component::GO:0033061': 1157, 'Cellular Component::GO:0043260': 1158, 'Cellular Component::GO:0000333': 1159, 'Cellular Component::GO:0071204': 1160, 'Cellular Component::GO:0005618': 1161, 'Cellular Component::GO:0044294': 1162, 'Cellular Component::GO:0005849': 1163, 'Cellular Component::GO:0048179': 1164, 'Cellular Component::GO:0002081': 1165, 'Cellular Component::GO:0043020': 1166, 'Cellular Component::GO:0033557': 1167, 'Cellular Component::GO:0045293': 1168, 'Cellular Component::GO:0001651': 1169, 'Cellular Component::GO:0036156': 1170, 'Cellular Component::GO:0097431': 1171, 'Cellular Component::GO:0016342': 1172, 'Cellular Component::GO:0031261': 1173, 'Cellular Component::GO:0005638': 1174, 'Cellular Component::GO:0009279': 1175, 'Cellular Component::GO:0005594': 1176, 'Cellular Component::GO:0000229': 1177, 'Cellular Component::GO:0030895': 1178, 'Cellular Component::GO:0008537': 1179, 'Cellular Component::GO:0032133': 1180, 'Cellular Component::GO:0002177': 1181, 'Cellular Component::GO:0009331': 1182, 'Cellular Component::GO:0030956': 1183, 'Cellular Component::GO:0019031': 1184, 'Cellular Component::GO:0031074': 1185, 'Cellular Component::GO:0090571': 1186, 'Cellular Component::GO:0071595': 1187, 'Cellular Component::GO:0014801': 1188, 'Cellular Component::GO:0000444': 1189, 'Cellular Component::GO:0097361': 1190, 'Cellular Component::GO:0001650': 1191, 'Cellular Component::GO:0030122': 1192, 'Cellular Component::GO:0042827': 1193, 'Cellular Component::GO:0034681': 1194, 'Cellular Component::GO:0031211': 1195, 'Cellular Component::GO:0008275': 1196, 'Cellular Component::GO:0005593': 1197, 'Cellular Component::GO:0019008': 1198, 'Cellular Component::GO:0036021': 1199, 'Cellular Component::GO:0070877': 1200, 'Cellular Component::GO:0030061': 1201, 'Cellular Component::GO:1990590': 1202, 'Cellular Component::GO:0044308': 1203, 'Cellular Component::GO:0044322': 1204, 'Cellular Component::GO:1990730': 1205, 'Cellular Component::GO:0030849': 1206, 'Cellular Component::GO:0005608': 1207, 'Cellular Component::GO:0031618': 1208, 'Cellular Component::GO:0098831': 1209, 'Cellular Component::GO:0032299': 1210, 'Cellular Component::GO:1990246': 1211, 'Cellular Component::GO:0032044': 1212, 'Cellular Component::GO:1990130': 1213, 'Cellular Component::GO:0034455': 1214, 'Cellular Component::GO:1990745': 1215, 'Cellular Component::GO:0005899': 1216, 'Cellular Component::GO:0034274': 1217, 'Cellular Component::GO:0031462': 1218, 'Cellular Component::GO:0070775': 1219, 'Cellular Component::GO:0032444': 1220, 'Cellular Component::GO:0005652': 1221, 'Cellular Component::GO:0036488': 1222, 'Cellular Component::GO:0030892': 1223, 'Cellular Component::GO:0014802': 1224, 'Cellular Component::GO:0000276': 1225, 'Cellular Component::GO:0042788': 1226, 'Cellular Component::GO:1990075': 1227, 'Cellular Component::GO:0045203': 1228, 'Cellular Component::GO:0033553': 1229, 'Cellular Component::GO:1903768': 1230, 'Cellular Component::GO:0097442': 1231, 'Cellular Component::GO:0044393': 1232, 'Cellular Component::GO:0061574': 1233, 'Cellular Component::GO:0097487': 1234, 'Cellular Component::GO:0031510': 1235, 'Cellular Component::GO:0005826': 1236, 'Cellular Component::GO:0070435': 1237, 'Cellular Component::GO:0033065': 1238, 'Cellular Component::GO:0030689': 1239, 'Cellular Component::GO:0030289': 1240, 'Cellular Component::GO:0005663': 1241, 'Cellular Component::GO:0034667': 1242, 'Cellular Component::GO:0005683': 1243, 'Cellular Component::GO:0033768': 1244, 'Cellular Component::GO:0070081': 1245, 'Cellular Component::GO:0097169': 1246, 'Cellular Component::GO:0044615': 1247, 'Cellular Component::GO:0034388': 1248, 'Cellular Component::GO:0018444': 1249, 'Cellular Component::GO:0034665': 1250, 'Cellular Component::GO:0071159': 1251, 'Cellular Component::GO:0060342': 1252, 'Cellular Component::GO:0031315': 1253, 'Cellular Component::GO:0097224': 1254, 'Cellular Component::GO:0048476': 1255, 'Cellular Component::GO:0000172': 1256, 'Cellular Component::GO:0000801': 1257, 'Cellular Component::GO:0071012': 1258, 'Cellular Component::GO:0000137': 1259, 'Cellular Component::GO:0070765': 1260, 'Cellular Component::GO:0097427': 1261, 'Cellular Component::GO:0032997': 1262, 'Cellular Component::GO:0010370': 1263, 'Cellular Component::GO:0033269': 1264, 'Cellular Component::GO:0016939': 1265, 'Cellular Component::GO:0001940': 1266, 'Cellular Component::GO:0071144': 1267, 'Cellular Component::GO:0005944': 1268, 'Cellular Component::GO:0002178': 1269, 'Cellular Component::GO:0034751': 1270, 'Cellular Component::GO:0033596': 1271, 'Cellular Component::GO:0043293': 1272, 'Cellular Component::GO:0005967': 1273, 'Cellular Component::GO:0016013': 1274, 'Cellular Component::GO:0030905': 1275, 'Cellular Component::GO:0044753': 1276, 'Cellular Component::GO:0038039': 1277, 'Cellular Component::GO:0031415': 1278, 'Cellular Component::GO:0002944': 1279, 'Cellular Component::GO:0031436': 1280, 'Cellular Component::GO:0048788': 1281, 'Cellular Component::GO:0031313': 1282, 'Cellular Component::GO:0042105': 1283, 'Cellular Component::GO:0035370': 1284, 'Cellular Component::GO:0045323': 1285, 'Cellular Component::GO:0009368': 1286, 'Cellular Component::GO:0000942': 1287, 'Cellular Component::GO:0035068': 1288, 'Cellular Component::GO:0034673': 1289, 'Cellular Component::GO:0017109': 1290, 'Cellular Component::GO:0061689': 1291, 'Cellular Component::GO:0020016': 1292, 'Cellular Component::GO:0016533': 1293, 'Cellular Component::GO:0030485': 1294, 'Cellular Component::GO:1990423': 1295, 'Cellular Component::GO:0033648': 1296, 'Cellular Component::GO:0002193': 1297, 'Cellular Component::GO:0060187': 1298, 'Cellular Component::GO:0097196': 1299, 'Cellular Component::GO:0035577': 1300, 'Cellular Component::GO:0031205': 1301, 'Cellular Component::GO:0072536': 1302, 'Cellular Component::GO:1990923': 1303, 'Cellular Component::GO:0030906': 1304, 'Cellular Component::GO:0031379': 1305, 'Cellular Component::GO:0070743': 1306, 'Cellular Component::GO:1990462': 1307, 'Cellular Component::GO:0070022': 1308, 'Cellular Component::GO:0060200': 1309, 'Cellular Component::GO:0035841': 1310, 'Cellular Component::GO:0030692': 1311, 'Cellular Component::GO:0046691': 1312, 'Cellular Component::GO:0005712': 1313, 'Cellular Component::GO:0048237': 1314, 'Cellular Component::GO:0032437': 1315, 'Cellular Component::GO:0042584': 1316, 'Cellular Component::GO:1990630': 1317, 'Cellular Component::GO:0071818': 1318, 'Cellular Component::GO:0034750': 1319, 'Cellular Component::GO:1990682': 1320, 'Cellular Component::GO:0032783': 1321, 'Cellular Component::GO:0005960': 1322, 'Cellular Component::GO:0045098': 1323, 'Cellular Component::GO:1990124': 1324, 'Cellular Component::GO:0034272': 1325, 'Cellular Component::GO:0036117': 1326, 'Cellular Component::GO:0030896': 1327, 'Cellular Component::GO:0016222': 1328, 'Cellular Component::GO:0070821': 1329, 'Cellular Component::GO:0032807': 1330, 'Cellular Component::GO:0032010': 1331, 'Cellular Component::GO:0031523': 1332, 'Cellular Component::GO:0043224': 1333, 'Cellular Component::GO:0005853': 1334, 'Cellular Component::GO:0090661': 1335, 'Cellular Component::GO:1990716': 1336, 'Cellular Component::GO:0033588': 1337, 'Cellular Component::GO:0097149': 1338, 'Cellular Component::GO:0002079': 1339, 'Cellular Component::GO:0032937': 1340, 'Cellular Component::GO:0034669': 1341, 'Cellular Component::GO:0090534': 1342, 'Cellular Component::GO:0072534': 1343, 'Cellular Component::GO:0097362': 1344, 'Cellular Component::GO:0089701': 1345, 'Cellular Component::GO:0000126': 1346, 'Cellular Component::GO:0070522': 1347, 'Cellular Component::GO:0071920': 1348, 'Cellular Component::GO:0070274': 1349, 'Cellular Component::GO:1990622': 1350, 'Cellular Component::GO:0002142': 1351, 'Cellular Component::GO:0043514': 1352, 'Cellular Component::GO:0044316': 1353, 'Cellular Component::GO:0042565': 1354, 'Cellular Component::GO:0097543': 1355, 'Cellular Component::GO:0036502': 1356, 'Cellular Component::GO:0060171': 1357, 'Cellular Component::GO:0032798': 1358, 'Cellular Component::GO:0005668': 1359, 'Cellular Component::GO:0060203': 1360, 'Cellular Component::GO:0032301': 1361, 'Cellular Component::GO:0044613': 1362, 'Cellular Component::GO:0038037': 1363, 'Cellular Component::GO:0031838': 1364, 'Cellular Component::GO:0097598': 1365, 'Cellular Component::GO:1990617': 1366, 'Cellular Component::GO:0036194': 1367, 'Cellular Component::GO:0035339': 1368, 'Cellular Component::GO:0016935': 1369, 'Cellular Component::GO:0035517': 1370, 'Cellular Component::GO:0043159': 1371, 'Cellular Component::GO:0031673': 1372, 'Cellular Component::GO:0005674': 1373, 'Cellular Component::GO:0072557': 1374, 'Cellular Component::GO:1990452': 1375, 'Cellular Component::GO:0098574': 1376, 'Cellular Component::GO:0044754': 1377, 'Cellular Component::GO:0005958': 1378, 'Cellular Component::GO:0034365': 1379, 'Cellular Component::GO:0070110': 1380, 'Cellular Component::GO:1990666': 1381, 'Cellular Component::GO:0061673': 1382, 'Cellular Component::GO:0031213': 1383, 'Cellular Component::GO:1990812': 1384, 'Cellular Component::GO:0070724': 1385, 'Cellular Component::GO:0071006': 1386, 'Cellular Component::GO:0005787': 1387, 'Cellular Component::GO:0031905': 1388, 'Cellular Component::GO:0005607': 1389, 'Cellular Component::GO:0009328': 1390}\n"
     ]
    }
   ],
   "source": [
    "entities_list = list(entity_dictionary.values())\n",
    "print(random.choice(entities_list)) #Imprime las entidades de un tipo específico de entidades. Cada entidad específica tiene a su vez un número ID particular."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "We0tfS7X-z9d",
   "metadata": {
    "id": "We0tfS7X-z9d"
   },
   "outputs": [],
   "source": [
    "#Encontramos los nodos y aristas involucrados en cada tipo de relación:\n",
    "edge_dictionary={} #Diccionario que contendra para cada tipo de relación, según una fuente de datos específica, una lista de tuplas\n",
    "                   #con los pares de nodos relacionados.\n",
    "node_dictionary={} #Diccionario que contendra para cada tipo de relación, según una fuente de datos específica, un conjunto de los IDs\n",
    "                   #de los nodos involucrados.\n",
    "for triple in triplets:\n",
    "    src = triple[0]\n",
    "    split_src=src.split('::')\n",
    "    src_type=split_src[0]\n",
    "    dest = triple[2]\n",
    "    split_dest=dest.split('::')\n",
    "    dest_type=split_dest[0]\n",
    "\n",
    "    src_int_id=entity_dictionary[src_type][src] #Guarda el ID de la entidad Head específica.\n",
    "    dest_int_id=entity_dictionary[dest_type][dest] # Idem para al entidad Tail relacionada con la anterior.\n",
    "\n",
    "    pair=[(src_int_id,dest_int_id)] #Genera una lista de tupla que contiene al par (id_head, id_tail) del triplete.\n",
    "    etype=triple[1] #da la relación entre las entidades del triplete\n",
    "    if etype in edge_dictionary:\n",
    "        edge_dictionary[etype]+=pair #incluye a la tupla de nodos que corresponden al tipo de relación correcto.\n",
    "    else:\n",
    "        edge_dictionary[etype]=pair\n",
    "\n",
    "    if etype in node_dictionary:\n",
    "        node_dictionary[etype].add(src_int_id) #método de adición de elementos a un conjunto.\n",
    "        node_dictionary[etype].add(dest_int_id)\n",
    "    else:\n",
    "        node_dictionary[etype]=set() #crea un conjunto, el cual es una colección desordenada y sin elementos duplicados.\n",
    "        node_dictionary[etype].add(src_int_id)\n",
    "        node_dictionary[etype].add(dest_int_id)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8Ez6K9CtBrBG",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1729024259729,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "8Ez6K9CtBrBG",
    "outputId": "a7f7639b-2770-4a76-fe8f-003181e077fa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('GNBR::X::Gene:Disease', {3, 33, 34, 35, 36, 24611, 39, 40, 41, 42, 24617, 45, 46, 48, 50, 55, 24631, 57, 58, 59, 24632, 62, 63, 66, 67, 68, 69, 71, 75, 8268, 78, 79, 24654, 16465, 83, 24660, 85, 87, 24667, 96, 99, 101, 8294, 104, 105, 107, 111, 118, 8311, 8312, 119, 120, 123, 24698, 8318, 127, 129, 130, 132, 140, 142, 146, 24723, 148, 149, 151, 152, 8344, 154, 8351, 8352, 8353, 162, 163, 159, 166, 24746, 177, 24755, 8377, 185, 8379, 194, 195, 197, 201, 203, 206, 209, 210, 214, 218, 219, 225, 228, 229, 233, 234, 236, 237, 238, 241, 242, 255, 262, 8455, 265, 272, 8473, 287, 8481, 292, 293, 295, 300, 301, 304, 305, 308, 8505, 313, 315, 316, 317, 24893, 319, 321, 322, 324, 24904, 329, 332, 24909, 16720, 338, 24917, 8536, 16741, 16744, 364, 365, 366, 367, 16750, 24940, 8565, 374, 377, 379, 380, 24955, 384, 386, 387, 388, 389, 392, 393, 394, 395, 396, 24973, 399, 411, 417, 420, 422, 8616, 425, 424, 430, 434, 439, 440, 441, 444, 450, 452, 25028, 460, 16846, 462, 25039, 469, 25046, 478, 479, 480, 483, 484, 25060, 8689, 498, 499, 500, 508, 509, 513, 515, 518, 8710, 25096, 523, 524, 526, 25105, 25107, 8726, 16921, 8741, 8747, 558, 8754, 566, 572, 576, 578, 580, 585, 586, 589, 596, 8788, 8789, 603, 610, 626, 8823, 634, 635, 636, 637, 638, 639, 8826, 643, 646, 647, 648, 649, 650, 651, 25228, 655, 17042, 660, 17045, 665, 8858, 669, 670, 671, 672, 17057, 674, 676, 679, 680, 25255, 682, 17067, 683, 684, 688, 8881, 708, 711, 713, 17099, 715, 723, 25306, 732, 25311, 25312, 8937, 753, 760, 761, 763, 771, 17163, 781, 25366, 800, 803, 812, 813, 816, 17206, 825, 828, 830, 9030, 841, 843, 845, 17230, 849, 857, 863, 865, 870, 871, 9063, 877, 9072, 881, 885, 25462, 9079, 888, 9080, 891, 9088, 897, 898, 900, 902, 903, 904, 907, 909, 912, 915, 25494, 17308, 927, 928, 931, 934, 936, 952, 954, 960, 963, 17347, 25553, 17365, 981, 995, 999, 17391, 1010, 17405, 1022, 1024, 1025, 1028, 17414, 9227, 1035, 9229, 1038, 9230, 1037, 9234, 1043, 1045, 17429, 1048, 1050, 1051, 1053, 1056, 17445, 1065, 9261, 1070, 9263, 9275, 1087, 1088, 25665, 9282, 1101, 1102, 17493, 17502, 1119, 25697, 1123, 1126, 1130, 1133, 1134, 25710, 1138, 1140, 17525, 1141, 1152, 17546, 25742, 1170, 1171, 1186, 1187, 1190, 1196, 1205, 25782, 1219, 1220, 1221, 1222, 1223, 1224, 1225, 1226, 17610, 1230, 1231, 9424, 1234, 1236, 1237, 1238, 1239, 1240, 1241, 25818, 1244, 1245, 1246, 1250, 1252, 1253, 1254, 9448, 9449, 1258, 17643, 1259, 1261, 25832, 9455, 1264, 1266, 9458, 25842, 1270, 1271, 1272, 1275, 1277, 17663, 1279, 17665, 1280, 1283, 1284, 1286, 1287, 1288, 1289, 1292, 1293, 1294, 1295, 1297, 1299, 1300, 1303, 9496, 1306, 1307, 1310, 1311, 17697, 1313, 25889, 1316, 1318, 25895, 1320, 1324, 1328, 1331, 25908, 1335, 1341, 1343, 25919, 25922, 25924, 1356, 1360, 1369, 25946, 1379, 1389, 1393, 1396, 1397, 1399, 17786, 1404, 1406, 1408, 1416, 1420, 1424, 1426, 1427, 1428, 26002, 1430, 17815, 1431, 1437, 1441, 1445, 1446, 17833, 17834, 1449, 1457, 1461, 1464, 1469, 1472, 1474, 26052, 1482, 1484, 1486, 1489, 9691, 17884, 1500, 1499, 26083, 1509, 1511, 1517, 1518, 17903, 17904, 1520, 1519, 1523, 9715, 1525, 26106, 1531, 1537, 1540, 26122, 1547, 17933, 1549, 9750, 1560, 1562, 1563, 9756, 1567, 1568, 26143, 1570, 1571, 17956, 1572, 1573, 26151, 26153, 1578, 17963, 1580, 1585, 1590, 1593, 1594, 1598, 9794, 1605, 26181, 26182, 1608, 1611, 1612, 9805, 1614, 1617, 9814, 1624, 26203, 18020, 1636, 18027, 26226, 26231, 1659, 1661, 26239, 1666, 1670, 18061, 18065, 1681, 26258, 1685, 1691, 1704, 1707, 26283, 26293, 1718, 26297, 18115, 1731, 1737, 26315, 1743, 26320, 1754, 9947, 1756, 1755, 9951, 9957, 1769, 9964, 9966, 9968, 18162, 1780, 18165, 18168, 1788, 1791, 1793, 1797, 1799, 1800, 1806, 1808, 1809, 18196, 1812, 1817, 1821, 1823, 1827, 18212, 18213, 1829, 1831, 10025, 10029, 1840, 1848, 10041, 1852, 18241, 1860, 1862, 1865, 1876, 1879, 10074, 18267, 1888, 1890, 18289, 1908, 10105, 1914, 1915, 18311, 1927, 18316, 18322, 1943, 1946, 18335, 18337, 1953, 1957, 10150, 10155, 18348, 18349, 1968, 1969, 1972, 10168, 1978, 1979, 1985, 1994, 1996, 2010, 10203, 2012, 2013, 2016, 2019, 18409, 10220, 2039, 10240, 2050, 2051, 10246, 2061, 2066, 2068, 18454, 2090, 10296, 2106, 2112, 2118, 2122, 2126, 10321, 2144, 2146, 18531, 2157, 2160, 18546, 2162, 2171, 18559, 18589, 2215, 2221, 2224, 10418, 18615, 2236, 2239, 2242, 10438, 18631, 2254, 2258, 2269, 2275, 2277, 18662, 2279, 2282, 18679, 10493, 2301, 2303, 2304, 2306, 2333, 2334, 2337, 2346, 2348, 2349, 2350, 18747, 18751, 2368, 2374, 2379, 10580, 2390, 2405, 10610, 10613, 18825, 2450, 10645, 10662, 18856, 18866, 10679, 10682, 2492, 18889, 10711, 18911, 2529, 18925, 18930, 18935, 18936, 2554, 10768, 18965, 10777, 2599, 10806, 19004, 10822, 10823, 19016, 10824, 10826, 19041, 10867, 2680, 2685, 19071, 19074, 2720, 10915, 2728, 10922, 2742, 2747, 10951, 10963, 10967, 2776, 10971, 10980, 10983, 10987, 10990, 19190, 2829, 2830, 11022, 11031, 11035, 2853, 2860, 19247, 19265, 11075, 2886, 19288, 19305, 11113, 11117, 11121, 2935, 19325, 19332, 11144, 2961, 19358, 19376, 2994, 3002, 19389, 11197, 19392, 3010, 11204, 3017, 3018, 3022, 19415, 3032, 3040, 3043, 19448, 3067, 3091, 3093, 19483, 19487, 19491, 19497, 3113, 3127, 11320, 3130, 11325, 3133, 19519, 3135, 3137, 19524, 3149, 19539, 3159, 3179, 3180, 3181, 19566, 11375, 3182, 3185, 3192, 19585, 3204, 3213, 3214, 3216, 3226, 3228, 3233, 3234, 3239, 11433, 11435, 3246, 11438, 3259, 3262, 11460, 3270, 3273, 3279, 11476, 3285, 3286, 3294, 3301, 19688, 19694, 19710, 3334, 3340, 11546, 11551, 11560, 3378, 3382, 11576, 11584, 11592, 3400, 11602, 3414, 3420, 19806, 19811, 3435, 11630, 3439, 11633, 11637, 3448, 11646, 3454, 3460, 11656, 3466, 11668, 19866, 11690, 3500, 11695, 11702, 3511, 19903, 11713, 19908, 3527, 3529, 3533, 3535, 11727, 3551, 3578, 3581, 3591, 3593, 3605, 3613, 3619, 3622, 3648, 11842, 3667, 11862, 3673, 11870, 20063, 3680, 3683, 20072, 3690, 20079, 11888, 11891, 20084, 3710, 20114, 20116, 20131, 11942, 3754, 20145, 11955, 3772, 20167, 3798, 3803, 12002, 12013, 3823, 3829, 20215, 3831, 3845, 3856, 3860, 20249, 3867, 3869, 3872, 20259, 3875, 3893, 20281, 3897, 12091, 3901, 3904, 3910, 3926, 3927, 3933, 3935, 3945, 20330, 3963, 20348, 3966, 3968, 3971, 3977, 20361, 3979, 3982, 3986, 3989, 12190, 4005, 20394, 20396, 4012, 4022, 4023, 12228, 4070, 4081, 4085, 20478, 4097, 4100, 12295, 4105, 12299, 12301, 12303, 12305, 4122, 12317, 4126, 4131, 20518, 12332, 4143, 12338, 12341, 12346, 12348, 12350, 4170, 4175, 20562, 12380, 4188, 4197, 4198, 4210, 4212, 4214, 12407, 12410, 12412, 4225, 4227, 20611, 12437, 20645, 4267, 4275, 4277, 12509, 4319, 4326, 4329, 12524, 4333, 12541, 4350, 12543, 4352, 12558, 20751, 12560, 20752, 4366, 4382, 12576, 20769, 12583, 4392, 4396, 4405, 12600, 12608, 12609, 12614, 12663, 20867, 20908, 12720, 20914, 4533, 12730, 4556, 20954, 20956, 12765, 21009, 12821, 12825, 12836, 12837, 12840, 4650, 12849, 4669, 12886, 12892, 12895, 12902, 12952, 12957, 12974, 21178, 21187, 21220, 13033, 21237, 13047, 13048, 21261, 13075, 13077, 21435, 21459, 21465, 21466, 21577, 21578, 13423, 21630, 13438, 21633, 21658, 21681, 13527, 13538, 21738, 13569, 13575, 21775, 21799, 21801, 21808, 13626, 21820, 13628, 21827, 21828, 21836, 13648, 13662, 21876, 21879, 13723, 21920, 21922, 21925, 21926, 13749, 13750, 13770, 13772, 13777, 13815, 13816, 22024, 13899, 22104, 22108, 22121, 13976, 13985, 22221, 22226, 22258, 22265, 22271, 22278, 14105, 22300, 22347, 22350, 22354, 14166, 22369, 22381, 22384, 14202, 14206, 22405, 14241, 14245, 14250, 14254, 14259, 14272, 22482, 6107, 22521, 22536, 22547, 14379, 14393, 14414, 14420, 14428, 14439, 22632, 22643, 14451, 22652, 22659, 22660, 22677, 22678, 22682, 22699, 22701, 22711, 22712, 14533, 22726, 14535, 14542, 22734, 22735, 14548, 14551, 22744, 14554, 14567, 14568, 22779, 22808, 22812, 14659, 22857, 14679, 14685, 22884, 22889, 22904, 22911, 22916, 6548, 6551, 22952, 22967, 22968, 6584, 22981, 6665, 6666, 6667, 6668, 6669, 6670, 14864, 6681, 14883, 6691, 6694, 23079, 6706, 6711, 6712, 23120, 6741, 6746, 23142, 23151, 6773, 23186, 6810, 6812, 6816, 6817, 6818, 6820, 6823, 6825, 6829, 6834, 6844, 6868, 6873, 6874, 23258, 15084, 15089, 6899, 6900, 23285, 15101, 15103, 23295, 23301, 6917, 6920, 23308, 23314, 6933, 6947, 6952, 6954, 6956, 15148, 6958, 6964, 6966, 6967, 23350, 6970, 6973, 6974, 23363, 6987, 6997, 23382, 7001, 15194, 23395, 7017, 7019, 7020, 7022, 7044, 7045, 7046, 15237, 7071, 7101, 7106, 7107, 7108, 23502, 7125, 7126, 7127, 7133, 23526, 23529, 15357, 23551, 7188, 23574, 23579, 23580, 23585, 7210, 7211, 7219, 23608, 7225, 7231, 23615, 7240, 23633, 23641, 15451, 23649, 15464, 7273, 15476, 7289, 7290, 7293, 15493, 7309, 23705, 15538, 23730, 15541, 7361, 7370, 7387, 7390, 15585, 7398, 23798, 15616, 23809, 7432, 23818, 23842, 7467, 15670, 7484, 7494, 7501, 7508, 7511, 7513, 7518, 7520, 7521, 23916, 7538, 15730, 7541, 7550, 7554, 7569, 23960, 7579, 7590, 7596, 7609, 24014, 7643, 15842, 7656, 24042, 24046, 7671, 7678, 24062, 24079, 24102, 7718, 7729, 7734, 24124, 7766, 7770, 7790, 7792, 24180, 15996, 24189, 7846, 24234, 24240, 7871, 7877, 24271, 7896, 7916, 7923, 7932, 24348, 24363, 24369, 7986, 7990, 16188, 7999, 16193, 24386, 8001, 8013, 8019, 8032, 8036, 24423, 24428, 8046, 8051, 8060, 8080, 24475, 8125, 24520, 16334, 24527, 8157, 16357, 24568})\n"
     ]
    }
   ],
   "source": [
    "nodes_list = list(node_dictionary.items())\n",
    "print(random.choice(nodes_list)) #Imprime para una relación de una fuente dada todos los ID de los nodos involucrados en dicha relación."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "BhgXTaGyBu4A",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 6,
     "status": "ok",
     "timestamp": 1729024259729,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "BhgXTaGyBu4A",
    "outputId": "5947ec11-292a-4e38-fe65-6ecb13b9a6c6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('DGIDB::ANTIBODY::Gene:Compound', [(3192, 6282), (156, 6351), (1482, 6409), (3192, 6416), (3260, 6485), (162, 6494), (119, 6662), (1389, 6799), (3359, 6813), (119, 6832), (3395, 6903), (843, 6915), (162, 6918), (127, 6945), (3421, 6315), (3196, 7018), (1788, 7026), (3442, 7043), (127, 6277), (1245, 6989), (3534, 7424), (3548, 7474), (393, 7480), (2775, 6441), (3220, 7550), (1877, 7585), (3259, 6478), (3589, 7598), (156, 7619), (3274, 6534), (162, 7688), (3628, 6356), (3640, 7742), (3652, 7785), (162, 7066), (3591, 7873), (3658, 7799), (1543, 7925), (162, 7972), (3437, 7034), (3395, 8046), (119, 6651), (3658, 6361), (483, 2755), (3348, 6777), (3561, 8263), (3196, 8312), (3796, 7474), (3801, 7873), (3492, 8368), (1306, 7475), (162, 8390), (3822, 8439), (3421, 8453), (3584, 8465), (3243, 8497), (393, 6912), (127, 6831), (110, 7289), (162, 8560), (3193, 8566), (127, 6370), (3584, 7602), (3196, 7455), (3872, 6989), (3861, 8539), (127, 8522), (115, 8720), (162, 8726), (1543, 6953), (3243, 8751), (3589, 8777), (1828, 6987), (403, 2375), (3196, 6309), (3391, 8750), (1958, 8897), (3058, 8939), (127, 6925), (3965, 8951), (110, 8978), (3225, 8998), (3196, 8399), (3196, 9001), (3464, 9020), (1245, 6273), (3421, 6789), (3196, 9005), (3243, 9117), (3266, 6500), (483, 6722), (483, 6016), (3609, 7649), (3348, 7342), (3381, 6953), (119, 9193), (3534, 9201), (127, 8497), (3196, 8450), (2787, 9320), (3962, 8941), (127, 9336), (2137, 9337), (1802, 9380), (1994, 9390), (3343, 9396), (3234, 6379), (3910, 7873), (1899, 9372), (3897, 9485), (127, 8890), (3533, 9337), (3869, 9547), (4117, 17), (578, 9560), (4066, 9294), (3217, 9598), (1788, 9561), (3439, 9442), (3534, 9605), (483, 8581), (3196, 9619), (3464, 9449), (3678, 9020), (156, 7475), (1039, 6086), (3496, 9168), (3348, 9594), (3930, 9708), (3348, 9736), (1501, 9782), (3534, 7799), (3259, 9830), (3439, 9259), (3179, 9337), (4203, 6990), (3410, 6808), (3806, 9881), (110, 10063), (3536, 6989), (3274, 6814), (3442, 10083), (3609, 10123), (3963, 4874), (3414, 10010), (111, 8969), (3201, 8368), (3628, 8439), (1973, 8720), (3930, 7873), (1754, 8540), (2775, 8750), (3220, 10258), (3658, 6808), (483, 8374), (3348, 9525), (3943, 10348), (3243, 8058), (3943, 10359), (3040, 6953), (483, 8433), (4293, 9341), (4049, 10245), (3437, 7342), (3328, 7873), (4303, 10441), (4243, 10113), (3395, 8699), (2853, 6858), (1543, 7047), (3243, 6820), (3910, 9708), (1994, 10115), (3348, 6921), (3040, 8270), (119, 10040), (1745, 8162), (3040, 7925), (1479, 9239), (3040, 7047), (3822, 6356), (483, 7684), (1973, 7638), (3438, 10709), (4378, 8270), (3658, 7450), (3439, 7036), (3590, 8270)])\n"
     ]
    }
   ],
   "source": [
    "edges_list = list(edge_dictionary.items())\n",
    "print(edges_list[14]) #Imprime para una relación de una fuente dada todos los pares de nodos conectados para dar ese tipo de relación."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bNo3mXNPb2cR",
   "metadata": {
    "id": "bNo3mXNPb2cR"
   },
   "source": [
    "#ENTRENAMIENTO DE KNOWLEDGE GRAPH EMBEDDINGS:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ckn70VFkgTTR",
   "metadata": {
    "id": "ckn70VFkgTTR"
   },
   "outputs": [],
   "source": [
    "nro_tripletes = len(df_drkg)\n",
    "seed = np.arange(nro_tripletes)\n",
    "np.random.shuffle(seed)\n",
    "\n",
    "train_cnt = int(nro_tripletes * 0.9) #Cantidad de datos para entrenamiento: 90% del total\n",
    "valid_cnt = int(nro_tripletes * 0.05) #Cantidad de datos para validación: 5% del total (igual a la cantidad para test)\n",
    "train_set = seed[:train_cnt]\n",
    "train_set = train_set.tolist()\n",
    "valid_set = seed[train_cnt:train_cnt+valid_cnt].tolist()\n",
    "test_set = seed[train_cnt+valid_cnt:].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6ehDgcUvgU_5",
   "metadata": {
    "id": "6ehDgcUvgU_5"
   },
   "outputs": [],
   "source": [
    "#Generación de archivos para cada set: train, validación y test. Se selecciona una carpeta donde guardar los mismos.\n",
    "\n",
    "with open(\"train_emb/drkg_train.tsv\", 'w+') as f:\n",
    "    for idx in train_set:\n",
    "        f.writelines(\"{}\\t{}\\t{}\\n\".format(triplets[idx][0], triplets[idx][1], triplets[idx][2]))\n",
    "\n",
    "with open(\"train_emb/drkg_valid.tsv\", 'w+') as f:\n",
    "    for idx in valid_set:\n",
    "        f.writelines(\"{}\\t{}\\t{}\\n\".format(triplets[idx][0], triplets[idx][1], triplets[idx][2]))\n",
    "\n",
    "with open(\"train_emb/drkg_test.tsv\", 'w+') as f:\n",
    "    for idx in test_set:\n",
    "        f.writelines(\"{}\\t{}\\t{}\\n\".format(triplets[idx][0], triplets[idx][1], triplets[idx][2]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "s4lEJkqKsEJe",
   "metadata": {
    "id": "s4lEJkqKsEJe"
   },
   "source": [
    "#El código de entrenamiento consta solo de una línea, utilizando la librería DGL-KE. Los autores entrenaron el modelo en una instancia de AWS p3.16xlarge de 8 GPUs.\n",
    "![aws gpus.jpg](data:image/jpeg;base64,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)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bydQaIT4s8bw",
   "metadata": {
    "id": "bydQaIT4s8bw"
   },
   "source": [
    "!DGLBACKEND=pytorch dglke_train --dataset DRKG --data_path ./train --data_files drkg_train.tsv drkg_valid.tsv drkg_test.tsv --format 'raw_udd_hrt' --model_name TransE_l2 --batch_size 2048 \\\n",
    "--neg_sample_size 256 --hidden_dim 400 --gamma 12.0 --lr 0.1 --max_step 100000 --log_interval 1000 --batch_size_eval 16 -adv --regularization_coef 1.00E-07 --test --num_thread 1 --gpu 0 1 2 3 4 5 6 7 --num_proc 8 --neg_sample_size_eval 10000 --async_update\n",
    "\n",
    "#No se correrá debido al requerimiento de equipo para realizar el entrenamiento.\n",
    "#El entrenamiento genera cuatro archivos:\n",
    "- Embeddings de entidades --> DRKG_TransE_l2_entity.npy\n",
    "- Embeddings de relaciones --> DRKG_TransE_l2_relation.npy\n",
    "- Mapeo de entidad a su ID --> entities.tsv\n",
    "- Mapeo de relación a su ID --> relations.tsv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "QhzquXEMs-IA",
   "metadata": {
    "id": "QhzquXEMs-IA"
   },
   "outputs": [],
   "source": [
    "#Embeddings de las entidades y de las relaciones:\n",
    "emb_entity = np.load('./embed/DRKG_TransE_l2_entity.npy') #Especificar la carpeta donde está guardado el archivo!\n",
    "emb_rel = np.load('./embed/DRKG_TransE_l2_relation.npy') #Especificar la carpeta donde está guardado el archivo!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "042XJcRjvJcu",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 477,
     "status": "ok",
     "timestamp": 1729193167601,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "042XJcRjvJcu",
    "outputId": "38487d4e-98fd-4aad-a645-9ebbb7c2ffb6"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97238, 400)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "emb_entity.shape #97238 embeddings de entidades de dimensión 400"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "G56pQAEjvRa8",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1729193167601,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "G56pQAEjvRa8",
    "outputId": "559dc5d2-1f18-4d7b-a215-1e5d6ee4326d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(107, 400)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "emb_rel.shape #107 embeddings de relaciones de dimensión 400"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Bx0D3jsXvphm",
   "metadata": {
    "id": "Bx0D3jsXvphm"
   },
   "outputs": [],
   "source": [
    "#entities.tsv, da el mapeo de entity_name a entity_id:\n",
    "columnas_ent = ['Entity_name','Entity_ID']\n",
    "df_entities = pd.read_csv('./embed/entities.tsv', sep='\\t', names = columnas_ent, header=None) #Especificar la carpeta donde está guardado el archivo!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "hGpo48lEvwrp",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 363
    },
    "executionInfo": {
     "elapsed": 480,
     "status": "ok",
     "timestamp": 1729193250523,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "hGpo48lEvwrp",
    "outputId": "edeb5c2d-d97a-4263-c40b-770a49db3064"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"df_entities\",\n  \"rows\": 10,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 10,\n        \"samples\": [\n          \"Symptom::D020233\",\n          \"Compound::pubchem:17754081\",\n          \"Compound::brenda:155472\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 30079,\n        \"min\": 3937,\n        \"max\": 84421,\n        \"num_unique_values\": 10,\n        \"samples\": [\n          82916,\n          8293,\n          3937\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-6148a617-ba41-4eb7-a97d-21fc0c0536b6\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>41800</th>\n",
       "      <td>Gene::6461</td>\n",
       "      <td>41800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8293</th>\n",
       "      <td>Compound::pubchem:17754081</td>\n",
       "      <td>8293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51404</th>\n",
       "      <td>Disease::MESH:C536385</td>\n",
       "      <td>51404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9596</th>\n",
       "      <td>Gene::3559</td>\n",
       "      <td>9596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10660</th>\n",
       "      <td>Compound::DB00681</td>\n",
       "      <td>10660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3937</th>\n",
       "      <td>Compound::brenda:155472</td>\n",
       "      <td>3937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21841</th>\n",
       "      <td>Gene::drugbank:BE0004305</td>\n",
       "      <td>21841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84421</th>\n",
       "      <td>Gene::57501</td>\n",
       "      <td>84421</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82916</th>\n",
       "      <td>Symptom::D020233</td>\n",
       "      <td>82916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34561</th>\n",
       "      <td>Gene::108701332</td>\n",
       "      <td>34561</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6148a617-ba41-4eb7-a97d-21fc0c0536b6')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-6148a617-ba41-4eb7-a97d-21fc0c0536b6 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-6148a617-ba41-4eb7-a97d-21fc0c0536b6');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-5dc4ccec-12f0-4403-ae30-6d8f933399ff\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-5dc4ccec-12f0-4403-ae30-6d8f933399ff')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-5dc4ccec-12f0-4403-ae30-6d8f933399ff button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                      Entity_name  Entity_ID\n",
       "41800                  Gene::6461      41800\n",
       "8293   Compound::pubchem:17754081       8293\n",
       "51404       Disease::MESH:C536385      51404\n",
       "9596                   Gene::3559       9596\n",
       "10660           Compound::DB00681      10660\n",
       "3937      Compound::brenda:155472       3937\n",
       "21841    Gene::drugbank:BE0004305      21841\n",
       "84421                 Gene::57501      84421\n",
       "82916            Symptom::D020233      82916\n",
       "34561             Gene::108701332      34561"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Selecciono 10 registros aleatoriamente:\n",
    "df_entities.sample(n=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "IvGEwA4EwLdi",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1030,
     "status": "ok",
     "timestamp": 1729193323736,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "IvGEwA4EwLdi",
    "outputId": "7f19aa32-3a5b-49c9-8b83-a8c05a8b1ccf"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(97238, 2)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_entities.shape #97238 entidades"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ubXrWh32v6_8",
   "metadata": {
    "id": "ubXrWh32v6_8"
   },
   "outputs": [],
   "source": [
    "#relations.tsv, da el mapeo de relation_name a relation_id:\n",
    "columnas_rel = ['Relation_name','Relation_ID']\n",
    "df_relations = pd.read_csv('./embed/relations.tsv', sep='\\t', names = columnas_rel, header=None) #Especificar la carpeta donde está guardado el archivo!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "mbZ0fVq1wdJX",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 363
    },
    "executionInfo": {
     "elapsed": 1039,
     "status": "ok",
     "timestamp": 1729193348430,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "mbZ0fVq1wdJX",
    "outputId": "fc91663a-7052-4004-e5a1-f33ee9c36ebf"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"df_relations\",\n  \"rows\": 10,\n  \"fields\": [\n    {\n      \"column\": \"Relation_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 10,\n        \"samples\": [\n          \"INTACT::ASSOCIATION::Gene:Gene\",\n          \"DGIDB::OTHER::Gene:Compound\",\n          \"GNBR::Sa::Compound:Disease\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Relation_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 35,\n        \"min\": 8,\n        \"max\": 98,\n        \"num_unique_values\": 10,\n        \"samples\": [\n          89,\n          8,\n          37\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-7b90dc57-baee-4825-a025-d4e95f8c8d32\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Relation_name</th>\n",
       "      <th>Relation_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>DGIDB::BINDER::Gene:Compound</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>DGIDB::OTHER::Gene:Compound</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>DRUGBANK::ddi-interactor-in::Compound:Compound</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>GNBR::E-::Compound:Gene</td>\n",
       "      <td>32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>DRUGBANK::target::Compound:Gene</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>GNBR::Sa::Compound:Disease</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>GNBR::O::Compound:Gene</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>INTACT::UBIQUITINATION REACTION::Gene:Gene</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>INTACT::ASSOCIATION::Gene:Gene</td>\n",
       "      <td>89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>INTACT::PROTEIN CLEAVAGE::Gene:Gene</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-7b90dc57-baee-4825-a025-d4e95f8c8d32')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-7b90dc57-baee-4825-a025-d4e95f8c8d32 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-7b90dc57-baee-4825-a025-d4e95f8c8d32');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-5f8b7a25-e9a2-4f6e-9699-5b62d0703c3c\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-5f8b7a25-e9a2-4f6e-9699-5b62d0703c3c')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-5f8b7a25-e9a2-4f6e-9699-5b62d0703c3c button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                                     Relation_name  Relation_ID\n",
       "10                    DGIDB::BINDER::Gene:Compound           10\n",
       "8                      DGIDB::OTHER::Gene:Compound            8\n",
       "20  DRUGBANK::ddi-interactor-in::Compound:Compound           20\n",
       "32                         GNBR::E-::Compound:Gene           32\n",
       "21                 DRUGBANK::target::Compound:Gene           21\n",
       "37                      GNBR::Sa::Compound:Disease           37\n",
       "33                          GNBR::O::Compound:Gene           33\n",
       "97      INTACT::UBIQUITINATION REACTION::Gene:Gene           97\n",
       "89                  INTACT::ASSOCIATION::Gene:Gene           89\n",
       "98             INTACT::PROTEIN CLEAVAGE::Gene:Gene           98"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Selecciono 10 registros aleatoriamente:\n",
    "df_relations.sample(n=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Yz9nYj6cwRcS",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 474,
     "status": "ok",
     "timestamp": 1729193355986,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "Yz9nYj6cwRcS",
    "outputId": "77ea533f-f8b3-42dd-99a8-d05741f0ffd1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(107, 2)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_relations.shape #97238 relaciones"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "AYsh2iIhwl3q",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 542,
     "status": "ok",
     "timestamp": 1729193780454,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "AYsh2iIhwl3q",
    "outputId": "44ee27e9-494a-4061-f77a-eb939c73feab"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of relations: 107\n"
     ]
    }
   ],
   "source": [
    "#Generación de 2 diccionarios de mapeo de relaciones a sus IDs y viceversa (IDs a relaciones):\n",
    "\n",
    "import csv\n",
    "\n",
    "rel2id = {}\n",
    "id2rel = {}\n",
    "with open(\"./embed/relations.tsv\", newline='', encoding='utf-8') as csvfile: #Especificar la carpeta donde está guardado el archivo!\n",
    "    reader = csv.DictReader(csvfile, delimiter='\\t', fieldnames=['rel','id'])\n",
    "    for row_val in reader:\n",
    "        id = row_val['id']\n",
    "        relation = row_val['rel']\n",
    "\n",
    "        rel2id[relation] = int(id)\n",
    "        id2rel[int(id)] = relation\n",
    "\n",
    "print(\"Number of relations: {}\".format(len(rel2id)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "BzZKmw_Pwlkq",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 447
    },
    "executionInfo": {
     "elapsed": 1582,
     "status": "ok",
     "timestamp": 1729193892809,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "BzZKmw_Pwlkq",
    "outputId": "3a313b8b-f38e-4ee0-b34a-bd9dcea3bc39"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.backends.backend_ps:The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Uso de t-SNE para convertir los embeddings de la relaciones en vectores de baja dimensión y así poder visualizar su distribución:\n",
    "from matplotlib import cm\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.utils import check_random_state\n",
    "from sklearn.manifold import TSNE\n",
    "\n",
    "dataset_id = {}\n",
    "for rel_name, i in rel2id.items():\n",
    "    rel_key = rel_name.split('::')[0]\n",
    "    if dataset_id.get(rel_key, None) is None:\n",
    "        dataset_id[rel_key] = []\n",
    "    dataset_id[rel_key].append(i)\n",
    "\n",
    "X_embedded = TSNE(n_components=2).fit_transform(emb_rel).T\n",
    "p = cm.rainbow(int(255/2 * 1))\n",
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111)\n",
    "for key, val in dataset_id.items():\n",
    "    val = np.asarray(val, dtype=np.longlong)\n",
    "\n",
    "    ax.plot(X_embedded[0][val], X_embedded[1][val], '.', label=key)\n",
    "\n",
    "lgd = ax.legend(bbox_to_anchor=(1.0, 1.0))\n",
    "plt.savefig('relation.eps', bbox_extra_artists=(lgd,), bbox_inches='tight', format='eps')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "jKbYm4G71gjq",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 522,
     "status": "ok",
     "timestamp": 1729194151537,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "jKbYm4G71gjq",
    "outputId": "04198077-3f8f-40f8-e1b9-c30308f9f3c8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of entities: 97238\n"
     ]
    }
   ],
   "source": [
    "#Generación de 2 diccionarios de mapeo de entidades a sus IDs y viceversa (IDs a entidades):\n",
    "entity2id = {}\n",
    "id2entity = {}\n",
    "with open(\"./embed/entities.tsv\", newline='', encoding='utf-8') as csvfile: #Especificar la carpeta donde está guardado el archivo!\n",
    "    reader = csv.DictReader(csvfile, delimiter='\\t', fieldnames=['entity','id'])\n",
    "    for row_val in reader:\n",
    "        id = row_val['id']\n",
    "        entity = row_val['entity']\n",
    "\n",
    "        entity2id[entity] = int(id)\n",
    "        id2entity[int(id)] = entity\n",
    "\n",
    "print(\"Number of entities: {}\".format(len(entity2id)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "BiFH9qKFJy3z",
   "metadata": {
    "id": "BiFH9qKFJy3z"
   },
   "outputs": [],
   "source": [
    "#Uso de t-SNE para convertir los embeddings de la entidades en vectores de baja dimensión y así poder visualizar su distribución:\n",
    "\n",
    "dataset_id = {}\n",
    "for entity_name, i in entity2id.items():\n",
    "    entity_key = entity_name.split('::')[0]\n",
    "    if dataset_id.get(entity_key, None) is None:\n",
    "        dataset_id[entity_key] = []\n",
    "    dataset_id[entity_key].append(i)\n",
    "\n",
    "X_embedded = TSNE(n_components=2).fit_transform(emb_entity).T #En el trabajo original de los autores se emplea una herramienta mucho más veloz para esta tarea.\n",
    "                                                              #Sin embargo, tras múltiples intentos, no lo pude hacer funcionar, por lo que opté por usar el paquete\n",
    "                                                              #empleado en el caso del t-SNE para las relaciones en el DRKG.\n",
    "p = cm.rainbow(int(255/2 * 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "og7bB2ESNOkk",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 447
    },
    "executionInfo": {
     "elapsed": 2158,
     "status": "ok",
     "timestamp": 1729197627155,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "og7bB2ESNOkk",
    "outputId": "4f5cf9a5-b7fb-4f8e-c246-07a4338f1749"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:matplotlib.backends.backend_ps:The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111)\n",
    "for key, val in dataset_id.items():\n",
    "    val = np.asarray(val, dtype=np.longlong)\n",
    "\n",
    "    ax.plot(X_embedded[0][val], X_embedded[1][val], '.', label=key)\n",
    "\n",
    "lgd = ax.legend(bbox_to_anchor=(1.0, 1.0))\n",
    "plt.savefig('entity.eps', bbox_extra_artists=(lgd,), bbox_inches='tight', format='eps')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4wYSFK5bPsCX",
   "metadata": {
    "id": "4wYSFK5bPsCX"
   },
   "source": [
    "#Drug Repurposing aplicado a las enfermedades autoinmunes desmielinizantes a través de relaciones enfermedad-compuestos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "KCmrtveESHe8",
   "metadata": {
    "id": "KCmrtveESHe8"
   },
   "outputs": [],
   "source": [
    "#Genero 3 dataframes: uno que contenga solo las entidades tipo Disease, otro las entidades tipo Drugs, y otro con entidades tipo Genes:\n",
    "n_ent = len(df_entities['Entity_name'])\n",
    "entidades_ = df_entities[df_entities['Entity_name'].str.startswith('Disease::')]\n",
    "entidades_drugs = df_entities[df_entities['Entity_name'].str.startswith('Compound::')]\n",
    "entidades_genes = df_entities[df_entities['Entity_name'].str.startswith('Gene::')]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Slm0QCTKj33V",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 489,
     "status": "ok",
     "timestamp": 1727995551026,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "Slm0QCTKj33V",
    "outputId": "4e882900-fbdc-421c-db3c-dc7092f83e52"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "24313"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(entidades_drugs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ilJx81rG0ALM",
   "metadata": {
    "id": "ilJx81rG0ALM"
   },
   "outputs": [],
   "source": [
    "drugs_in_drkg = (entidades_drugs['Entity_name']).tolist() #Lista de todas las drogas en el DRKG"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Lqr7vOTWVM6W",
   "metadata": {
    "id": "Lqr7vOTWVM6W"
   },
   "source": [
    "**Para la tarea de reposicionamiento de medicamentos, se eligieron las enfermedades autoinmunes desmielinizantes.En la web https://www.nlm.nih.gov/se hallarán sus ID MESH y luego se verificará si están en el DRKG.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Ys1CWgbEYDMJ",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 680,
     "status": "ok",
     "timestamp": 1729198323293,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "Ys1CWgbEYDMJ",
    "outputId": "4c582354-049f-4dbf-e132-3a877162368a"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"entidades_[entidades_['Entity_name'] == 'Disease::MESH:D020278']\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Disease::MESH:D020278\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 47285,\n        \"max\": 47285,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          47285\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-52fae25a-dc34-4dd4-89d7-cee57458c7bd\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>47285</th>\n",
       "      <td>Disease::MESH:D020278</td>\n",
       "      <td>47285</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-52fae25a-dc34-4dd4-89d7-cee57458c7bd')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-52fae25a-dc34-4dd4-89d7-cee57458c7bd button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-52fae25a-dc34-4dd4-89d7-cee57458c7bd');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                 Entity_name  Entity_ID\n",
       "47285  Disease::MESH:D020278      47285"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Enfermedades autoinmunes desmielinizantes (EADs): ID MESH D020278\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D020278'] #Verifico su presencia en el DRKG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "mbOEWMY6Xe7z",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 755,
     "status": "ok",
     "timestamp": 1729198336729,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "mbOEWMY6Xe7z",
    "outputId": "b6273882-4e0b-4b2d-c11b-6bfd5f8facac"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"entidades_[entidades_['Entity_name'] == 'Disease::MESH:D009103'] #Verifico su presencia en el DRKG\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Disease::MESH:D009103\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 25719,\n        \"max\": 25719,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          25719\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-f9a4f978-5a28-4b78-92be-9102f74b975b\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>25719</th>\n",
       "      <td>Disease::MESH:D009103</td>\n",
       "      <td>25719</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f9a4f978-5a28-4b78-92be-9102f74b975b')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-f9a4f978-5a28-4b78-92be-9102f74b975b button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-f9a4f978-5a28-4b78-92be-9102f74b975b');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                 Entity_name  Entity_ID\n",
       "25719  Disease::MESH:D009103      25719"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Esclerosis mútiple: ID MESH D009103\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D009103'] #Verifico su presencia en el DRKG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "M3_pIn36dJ1b",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 697,
     "status": "ok",
     "timestamp": 1729198377241,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "M3_pIn36dJ1b",
    "outputId": "33488320-6d3a-48f9-bdb5-6023c8280909"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"entidades_[entidades_['Entity_name'] == 'Disease::MESH:D020529'] #Verifico su presencia en el DRKG\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Disease::MESH:D020529\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 25152,\n        \"max\": 25152,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          25152\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-0b69061b-13da-43d1-8ff7-59c36d0df396\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>25152</th>\n",
       "      <td>Disease::MESH:D020529</td>\n",
       "      <td>25152</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0b69061b-13da-43d1-8ff7-59c36d0df396')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-0b69061b-13da-43d1-8ff7-59c36d0df396 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-0b69061b-13da-43d1-8ff7-59c36d0df396');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                 Entity_name  Entity_ID\n",
       "25152  Disease::MESH:D020529      25152"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Esclerosis múltiple recurrente-remitente: ID MESH D020529\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D020529'] #Verifico su presencia en el DRKG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Hle4caOxctpt",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 790,
     "status": "ok",
     "timestamp": 1729198410059,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "Hle4caOxctpt",
    "outputId": "684b06af-e07e-4963-ead9-424b3b0f0aa3"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"entidades_[entidades_['Entity_name'] == 'Disease::MESH:D020528'] #Verifico su presencia en el DRKG\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Disease::MESH:D020528\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 25163,\n        \"max\": 25163,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          25163\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-93b67934-a6eb-4e6c-81bf-25fdbd017455\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>25163</th>\n",
       "      <td>Disease::MESH:D020528</td>\n",
       "      <td>25163</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-93b67934-a6eb-4e6c-81bf-25fdbd017455')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-93b67934-a6eb-4e6c-81bf-25fdbd017455 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-93b67934-a6eb-4e6c-81bf-25fdbd017455');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                 Entity_name  Entity_ID\n",
       "25163  Disease::MESH:D020528      25163"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Esclerosis múltiple progresiva crónica: ID MESH D020528\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D020528'] #Verifico su presencia en el DRKG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "lu8ysqrzgYWD",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 773,
     "status": "ok",
     "timestamp": 1729198442097,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "lu8ysqrzgYWD",
    "outputId": "7aa512eb-cc02-4807-a248-66d31424418d"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"entidades_[entidades_['Entity_name'] == 'Disease::MESH:D009471']\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Disease::MESH:D009471\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 26072,\n        \"max\": 26072,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          26072\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-4ad4eef1-27ac-48da-a3f6-13d2ff7a97b6\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>26072</th>\n",
       "      <td>Disease::MESH:D009471</td>\n",
       "      <td>26072</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-4ad4eef1-27ac-48da-a3f6-13d2ff7a97b6')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-4ad4eef1-27ac-48da-a3f6-13d2ff7a97b6 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-4ad4eef1-27ac-48da-a3f6-13d2ff7a97b6');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                 Entity_name  Entity_ID\n",
       "26072  Disease::MESH:D009471      26072"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Neuromielitis óptica: ID MESH D009471\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D009471'] #Verifico su presencia en el DRKG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "KO5cePergDhm",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 1030,
     "status": "ok",
     "timestamp": 1729198461181,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "KO5cePergDhm",
    "outputId": "4dd3d1e8-45b1-4b17-9f30-1ca2811c94a5"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"entidades_[entidades_['Entity_name'] == 'Disease::MESH:D009188']\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Disease::MESH:D009188\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 49257,\n        \"max\": 49257,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          49257\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-b1e89934-01f7-4455-9943-133442805824\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>49257</th>\n",
       "      <td>Disease::MESH:D009188</td>\n",
       "      <td>49257</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b1e89934-01f7-4455-9943-133442805824')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-b1e89934-01f7-4455-9943-133442805824 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-b1e89934-01f7-4455-9943-133442805824');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                 Entity_name  Entity_ID\n",
       "49257  Disease::MESH:D009188      49257"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Mielitis transversa: ID MESH D009188\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D009188'] #Verifico su presencia en el DRKG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "KS0DkazId_kJ",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 1023,
     "status": "ok",
     "timestamp": 1729198537123,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "KS0DkazId_kJ",
    "outputId": "5a69e420-f77a-4232-941b-edd3d199cc0a"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"entidades_[entidades_['Entity_name'] == 'Disease::MESH:D002549'] #Verifico su presencia en el DRKG\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Disease::MESH:D002549\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 56534,\n        \"max\": 56534,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          56534\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-1b5ff7b0-a776-409a-8709-bbfa7d52ab04\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>56534</th>\n",
       "      <td>Disease::MESH:D002549</td>\n",
       "      <td>56534</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1b5ff7b0-a776-409a-8709-bbfa7d52ab04')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-1b5ff7b0-a776-409a-8709-bbfa7d52ab04 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-1b5ff7b0-a776-409a-8709-bbfa7d52ab04');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                 Entity_name  Entity_ID\n",
       "56534  Disease::MESH:D002549      56534"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Esclerosis cerebral difusa de Schilder (esclerosis concéntrica de Balo): ID MESH D002549\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D002549'] #Verifico su presencia en el DRKG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "vP4lfCVKkOT3",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 1031,
     "status": "ok",
     "timestamp": 1729198641295,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "vP4lfCVKkOT3",
    "outputId": "a1842c43-a825-4250-d26b-1f2af9431798"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"entidades_[entidades_['Entity_name'] == 'Disease::MESH:D004673'] # Verifico su presencia en el DRKG, pero no la trendr\\u00e9 en cuenta!\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"Entity_name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Disease::MESH:D004673\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Entity_ID\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 46631,\n        \"max\": 46631,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          46631\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-bc7be725-f90d-4b88-ac90-1cad9549b32e\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>46631</th>\n",
       "      <td>Disease::MESH:D004673</td>\n",
       "      <td>46631</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-bc7be725-f90d-4b88-ac90-1cad9549b32e')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-bc7be725-f90d-4b88-ac90-1cad9549b32e button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-bc7be725-f90d-4b88-ac90-1cad9549b32e');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                 Entity_name  Entity_ID\n",
       "46631  Disease::MESH:D004673      46631"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Encefalomielitis aguda diseminada: ID MESH D004673  ----------> Enfermedad monofásica, no hay drogas que puedan modificarla (surge de un evento, no tiene una evolución como las otras).\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D004673'] # Verifico su presencia en el DRKG, pero no la trendré en cuenta!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "TCu3GU2_bmF4",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 53
    },
    "executionInfo": {
     "elapsed": 1024,
     "status": "ok",
     "timestamp": 1729198654051,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "TCu3GU2_bmF4",
    "outputId": "a60b6df2-d6b1-449c-baef-42a9c8acab08"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "repr_error": "Out of range float values are not JSON compliant: nan",
       "type": "dataframe"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-7b33a289-9085-4588-afb4-b245b8ae0c0e\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Entity_name</th>\n",
       "      <th>Entity_ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-7b33a289-9085-4588-afb4-b245b8ae0c0e')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-7b33a289-9085-4588-afb4-b245b8ae0c0e button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-7b33a289-9085-4588-afb4-b245b8ae0c0e');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [Entity_name, Entity_ID]\n",
       "Index: []"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Leucoencefalitis aguda hemorrágica: D004684 ----------> Enfermedad monofásica, no hay drogas que puedan modificarla (surge de un evento, no tiene una evolución como las otras).\n",
    "entidades_[entidades_['Entity_name'] == 'Disease::MESH:D004684'] # Verifico su presencia en el DRKG, pero no la trendré en cuenta!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "mAu7QGZRhGj-",
   "metadata": {
    "id": "mAu7QGZRhGj-"
   },
   "outputs": [],
   "source": [
    "#Hago una lista con los IDs de las enfermedades autoinmunes desmielinizantes (EADs) encontradas y que se pueden tratar:\n",
    "lista_EAD = [\n",
    "'Disease::MESH:D009103',\n",
    "'Disease::MESH:D020278',\n",
    "'Disease::MESH:D002549',\n",
    "'Disease::MESH:D020528',\n",
    "'Disease::MESH:D020529',\n",
    "'Disease::MESH:D009188',\n",
    "'Disease::MESH:D009471'\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "kV_ok7cGQw9v",
   "metadata": {
    "id": "kV_ok7cGQw9v"
   },
   "outputs": [],
   "source": [
    "# Usamos como drogas candidatas aquellas aprobadas por la FDA que se encuentran en Drugbank.\n",
    "# (se excluyen drogas con peso molecular menor a 250). Esta lista de drogas está en el archivo infer_drug.tsv.\n",
    "\n",
    "drug_list = []\n",
    "with open(\"./infer_drug.tsv\", newline='', encoding='utf-8') as csvfile: #Especificar la carpeta donde está guardado el archivo!\n",
    "    reader = csv.DictReader(csvfile, delimiter='\\t', fieldnames=['drug','ids'])\n",
    "    for row_val in reader:\n",
    "        drug_list.append(row_val['drug'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "BCIUa0WxSKnP",
   "metadata": {
    "id": "BCIUa0WxSKnP"
   },
   "outputs": [],
   "source": [
    "entity_idmap_file = './embed/entities.tsv' #columnas: entidad (por ej, Gene::864) y su ID numérico único (12162).\n",
    "relation_idmap_file = './embed/relations.tsv' #columnas: relación (GNBR::T::Compound:Disease) y su ID numérico único (35)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "o2u0NPSsUkMt",
   "metadata": {
    "id": "o2u0NPSsUkMt"
   },
   "outputs": [],
   "source": [
    "# Generamos una lista con los ID únicos para las enfermedades desmielinizantes, otra con los IDs de las drogas y otra con los de los tratamientos elegidos del mapa completo de relaciones.\n",
    "entity_map = {}\n",
    "entity_id_map = {}\n",
    "relation_map = {}\n",
    "with open(entity_idmap_file, newline='', encoding='utf-8') as csvfile:\n",
    "    reader = csv.DictReader(csvfile, delimiter='\\t', fieldnames=['name','id'])\n",
    "    for row_val in reader:\n",
    "        entity_map[row_val['name']] = int(row_val['id'])\n",
    "        entity_id_map[int(row_val['id'])] = row_val['name']\n",
    "\n",
    "with open(relation_idmap_file, newline='', encoding='utf-8') as csvfile:\n",
    "    reader = csv.DictReader(csvfile, delimiter='\\t', fieldnames=['name','id'])\n",
    "    for row_val in reader:\n",
    "        relation_map[row_val['name']] = int(row_val['id'])\n",
    "\n",
    "# handle the ID mapping\n",
    "drug_ids = []\n",
    "disease_ids_EAD = []\n",
    "for drug in drug_list:\n",
    "    drug_ids.append(entity_map[drug])\n",
    "\n",
    "for disease in lista_EAD:\n",
    "    disease_ids_EAD.append(entity_map[disease])\n",
    "\n",
    "# Tenemos tres relaciones de Tratamiento:\n",
    "treatment = ['DRUGBANK::treats::Compound:Disease','Hetionet::CtD::Compound:Disease','GNBR::T::Compound:Disease']\n",
    "treatment_rid_EAD = [relation_map[treat]  for treat in treatment]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "OGsK141mjOoB",
   "metadata": {
    "id": "OGsK141mjOoB"
   },
   "outputs": [],
   "source": [
    "# Cargamos los embeddings de los tratamientos relacionados a las EAD:\n",
    "drug_ids = torch.tensor(drug_ids).long()\n",
    "disease_ids_EAD = torch.tensor(disease_ids_EAD).long()\n",
    "treatment_rid_EAD = torch.tensor(treatment_rid_EAD)\n",
    "\n",
    "drug_emb = torch.tensor(emb_entity[drug_ids])\n",
    "treatment_embs_EAD = [torch.tensor(emb_rel[rid]) for rid in treatment_rid_EAD]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2VnqNuNRkXx7",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 317,
     "status": "ok",
     "timestamp": 1718397473307,
     "user": {
      "displayName": "Nicolás Eloy Barrera",
      "userId": "12716743868083471095"
     },
     "user_tz": 180
    },
    "id": "2VnqNuNRkXx7",
    "outputId": "9972d429-f665-4509-8366-b14ed8fbbba7"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "400"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(treatment_embs_EAD[0]) #cada embedding consta de 400 dimensiones."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "A66etzMaYQhN",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 757,
     "status": "ok",
     "timestamp": 1729200476578,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "A66etzMaYQhN",
    "outputId": "864172b5-bbdf-4999-fe1e-d01e22df8159"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "400"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(drug_emb[0]) #cada embedding consta de 400 dimensiones."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "-NzN2hC_WIVR",
   "metadata": {
    "id": "-NzN2hC_WIVR"
   },
   "source": [
    "Usamos el modelo TransE de embeddings con un score tipo log-sigmoide para hacer todos los scores negativos.\n",
    "\n",
    "$\\mathbf{d} = \\gamma - ||\\mathbf{h}+\\mathbf{r}-\\mathbf{t}||_{2}$\n",
    "\n",
    "$\\mathbf{score} = \\log\\left(\\frac{1}{1+\\exp(\\mathbf{-d})}\\right)$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "tZzSO0SPlCWN",
   "metadata": {
    "id": "tZzSO0SPlCWN"
   },
   "outputs": [],
   "source": [
    "import torch.nn.functional as fn\n",
    "\n",
    "gamma=12.0 #Hiperparametro que se fija mediante Grid Search.\n",
    "\n",
    "def transE_l2(head, rel, tail):\n",
    "    score = head + rel - tail\n",
    "    return gamma - torch.norm(score, p=2, dim=-1)\n",
    "\n",
    "scores_per_disease_EAD = []\n",
    "dids_EAD = []\n",
    "for rid in range(len(treatment_embs_EAD)):\n",
    "    treatment_emb_EAD=treatment_embs_EAD[rid]\n",
    "    for disease_id in disease_ids_EAD:\n",
    "        disease_emb_EAD = emb_entity[disease_id]\n",
    "        score_EAD = fn.logsigmoid(transE_l2(drug_emb, treatment_emb_EAD, disease_emb_EAD))\n",
    "        scores_per_disease_EAD.append(score_EAD)\n",
    "        dids_EAD.append(drug_ids)\n",
    "scores_EAD = torch.cat(scores_per_disease_EAD)\n",
    "dids_EAD = torch.cat(dids_EAD)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Vkz_EWDhl5ay",
   "metadata": {
    "id": "Vkz_EWDhl5ay"
   },
   "outputs": [],
   "source": [
    "# Ordenamos los scores en orden descendente\n",
    "idx = torch.flip(torch.argsort(scores_EAD), dims=[0])\n",
    "scores_EAD = scores_EAD[idx].numpy()\n",
    "dids_EAD = dids_EAD[idx].numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "KliOMnHpmlcF",
   "metadata": {
    "id": "KliOMnHpmlcF"
   },
   "outputs": [],
   "source": [
    "#Hallamos los tratamientos propuestos:\n",
    "_, unique_indices_EAD = np.unique(dids_EAD, return_index=True)\n",
    "topk_EAD=100\n",
    "topk_indices_EAD = np.sort(unique_indices_EAD)[:topk_EAD]\n",
    "proposed_dids_EAD = dids_EAD[topk_indices_EAD]\n",
    "proposed_scores_EAD = scores_EAD[topk_indices_EAD]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "l1XxbjEKdEHR",
   "metadata": {
    "id": "l1XxbjEKdEHR"
   },
   "outputs": [],
   "source": [
    "#drugbank.tsv contiene la información listada en el .xml de la base de datos de DrugBank (no son todas las drogas,\n",
    "#por lo que puede haber algunas sin identificar y se deberán buscar a mano en la página de DrugBank):\n",
    "df_drugbank = pd.read_csv('drugbank.tsv', sep='\\t') #Especificar la carpeta donde está guardado el archivo!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "15Tmmoz_eW6z",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 1041,
     "status": "ok",
     "timestamp": 1729200649660,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "15Tmmoz_eW6z",
    "outputId": "bd1e7dd3-555e-495e-fead-82b9a79fb32a"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"df_drugbank\",\n  \"rows\": 7759,\n  \"fields\": [\n    {\n      \"column\": \"drugbank_id\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 7759,\n        \"samples\": [\n          \"DB02249\",\n          \"DB03868\",\n          \"DB01362\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"name\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 7759,\n        \"samples\": [\n          \"2-Ethoxyethanol\",\n          \"3-Dehydroquinic Acid\",\n          \"Iohexol\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"type\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"small molecule\",\n          \"biotech\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"groups\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 27,\n        \"samples\": [\n          \"approved|illicit\",\n          \"investigational|withdrawn\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"atc_codes\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 1480,\n        \"samples\": [\n          \"L01XX19\",\n          \"A02BC01|A02BC05\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"categories\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 953,\n        \"samples\": [\n          \"Antihypermagnesemic Agents|Antihypocalcemic Agents|Electrolyte replenisher|Supplements\",\n          \"Analgesics|Anticonvulsants|Parasympatholytics|Muscle Relaxants, Central|Adrenergic alpha-2 Receptor Agonists|Neuromuscular Agents|Muscle Relaxants, Skeletal\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"inchikey\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 6682,\n        \"samples\": [\n          \"InChIKey=MITFXPHMIHQXPI-UHFFFAOYSA-N\",\n          \"InChIKey=BIHLSRJPJFOESJ-UHFFFAOYSA-N\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"inchi\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 6682,\n        \"samples\": [\n          \"InChI=1S/C16H12ClNO3/c1-9(16(19)20)11-4-7-14-13(8-11)18-15(21-14)10-2-5-12(17)6-3-10/h2-9H,1H3,(H,19,20)\",\n          \"InChI=1S/C15H13NO/c1-10-7-11(2)9-12(8-10)15-16-13-5-3-4-6-14(13)17-15/h3-9H,1-2H3\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"description\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3043,\n        \"samples\": [\n          \"Bexarotene (Targretin) is an antineoplastic agent indicated by the FDA for Cutaneous T cell lymphoma. It has been used off-label for lung cancer, breast cancer, and Kaposi's sarcoma. [Wikipedia]\",\n          \"Ginsenosides are a class of steroid glycosides, and triterpene saponins, found exclusively in the plant genus Panax (ginseng). Ginsenosides have been the target of research, as they are viewed as the active compounds behind the claims of ginseng's efficacy. Because ginsenosides appear to affect multiple pathways, their effects are complex and difficult to isolate. Rg1 Appears to be most abundant in Panax ginseng (Chinese/Korean Ginseng). It improves spatial learning and increase hippocampal synaptophysin level in mice, plus demonstrates estrogen-like activity.\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe",
       "variable_name": "df_drugbank"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-0d34a3c2-1174-4a32-a1c6-3f9f4f02b732\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>drugbank_id</th>\n",
       "      <th>name</th>\n",
       "      <th>type</th>\n",
       "      <th>groups</th>\n",
       "      <th>atc_codes</th>\n",
       "      <th>categories</th>\n",
       "      <th>inchikey</th>\n",
       "      <th>inchi</th>\n",
       "      <th>description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>DB00001</td>\n",
       "      <td>Lepirudin</td>\n",
       "      <td>biotech</td>\n",
       "      <td>approved</td>\n",
       "      <td>B01AE02</td>\n",
       "      <td>Antithrombins|Fibrinolytic Agents</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Lepirudin is identical to natural hirudin exce...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DB00002</td>\n",
       "      <td>Cetuximab</td>\n",
       "      <td>biotech</td>\n",
       "      <td>approved</td>\n",
       "      <td>L01XC06</td>\n",
       "      <td>Antineoplastic Agents</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Epidermal growth factor receptor binding FAB. ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>DB00003</td>\n",
       "      <td>Dornase alfa</td>\n",
       "      <td>biotech</td>\n",
       "      <td>approved</td>\n",
       "      <td>R05CB13</td>\n",
       "      <td>Enzymes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Dornase alfa is a biosynthetic form of human d...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>DB00004</td>\n",
       "      <td>Denileukin diftitox</td>\n",
       "      <td>biotech</td>\n",
       "      <td>approved|investigational</td>\n",
       "      <td>L01XX29</td>\n",
       "      <td>Antineoplastic Agents</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>A recombinant DNA-derived cytotoxic protein co...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>DB00005</td>\n",
       "      <td>Etanercept</td>\n",
       "      <td>biotech</td>\n",
       "      <td>approved|investigational</td>\n",
       "      <td>L04AB01</td>\n",
       "      <td>Immunosuppressive Agents</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Dimeric fusion protein consisting of the extra...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0d34a3c2-1174-4a32-a1c6-3f9f4f02b732')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-0d34a3c2-1174-4a32-a1c6-3f9f4f02b732 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-0d34a3c2-1174-4a32-a1c6-3f9f4f02b732');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-9d566e2b-2ee5-411a-a82a-ffce53c60561\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-9d566e2b-2ee5-411a-a82a-ffce53c60561')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-9d566e2b-2ee5-411a-a82a-ffce53c60561 button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "  drugbank_id                 name     type                    groups  \\\n",
       "0     DB00001            Lepirudin  biotech                  approved   \n",
       "1     DB00002            Cetuximab  biotech                  approved   \n",
       "2     DB00003         Dornase alfa  biotech                  approved   \n",
       "3     DB00004  Denileukin diftitox  biotech  approved|investigational   \n",
       "4     DB00005           Etanercept  biotech  approved|investigational   \n",
       "\n",
       "  atc_codes                         categories inchikey inchi  \\\n",
       "0   B01AE02  Antithrombins|Fibrinolytic Agents      NaN   NaN   \n",
       "1   L01XC06              Antineoplastic Agents      NaN   NaN   \n",
       "2   R05CB13                            Enzymes      NaN   NaN   \n",
       "3   L01XX29              Antineoplastic Agents      NaN   NaN   \n",
       "4   L04AB01           Immunosuppressive Agents      NaN   NaN   \n",
       "\n",
       "                                         description  \n",
       "0  Lepirudin is identical to natural hirudin exce...  \n",
       "1  Epidermal growth factor receptor binding FAB. ...  \n",
       "2  Dornase alfa is a biosynthetic form of human d...  \n",
       "3  A recombinant DNA-derived cytotoxic protein co...  \n",
       "4  Dimeric fusion protein consisting of the extra...  "
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_drugbank.head() #primeros registros en el DataFrame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "LEEt_LiPkFpb",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 688,
     "status": "ok",
     "timestamp": 1729200702493,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "LEEt_LiPkFpb",
    "outputId": "a5441f23-72e8-44a1-ab03-5d7470f1efc0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7759"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_drugbank) #Cantidad de registros en el DataFrame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "IgJQIbmSn9ZW",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 487,
     "status": "ok",
     "timestamp": 1729200604106,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "IgJQIbmSn9ZW",
    "outputId": "a5b30acb-e047-45fa-c0a8-e332ffa289b2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\tMethylprednisolone\t-0.12100826948881149\n",
      "[1]\tAzathioprine\t-0.12112560868263245\n",
      "[2]\tMethotrexate\t-0.12847775220870972\n",
      "[3]\tDexamethasone\t-0.12967625260353088\n",
      "[4]\tMelatonin\t-0.132446750998497\n",
      "[5]\tTeriflunomide\t-0.13346567749977112\n",
      "[6]\tDB11094\t-0.1414821892976761\n",
      "[7]\tCladribine\t-0.14233370125293732\n",
      "[8]\tThalidomide\t-0.1431959718465805\n",
      "[9]\tEstradiol\t-0.14724911749362946\n",
      "[10]\tDB11672\t-0.1509384959936142\n",
      "[11]\tDB14783\t-0.15210206806659698\n",
      "[12]\tCyclosporine\t-0.15574337542057037\n",
      "[13]\tCholesterol\t-0.15823456645011902\n",
      "[14]\tHydrocortisone\t-0.16551189124584198\n",
      "[15]\tDB13038\t-0.1685316413640976\n",
      "[16]\tMorphine\t-0.1695857048034668\n",
      "[17]\tPrednisone\t-0.16970013082027435\n",
      "[18]\tSimvastatin\t-0.17194552719593048\n",
      "[19]\tIndomethacin\t-0.1736704409122467\n",
      "[20]\tFluticasone Propionate\t-0.17437157034873962\n",
      "[21]\tColchicine\t-0.1745726466178894\n",
      "[22]\tDoxorubicin\t-0.17502200603485107\n",
      "[23]\tMitoxantrone\t-0.1756279468536377\n",
      "[24]\tTretinoin\t-0.1766490787267685\n",
      "[25]\tPentoxifylline\t-0.17967404425144196\n",
      "[26]\tCisplatin\t-0.18140099942684174\n",
      "[27]\tInterferon beta-1b\t-0.1818210333585739\n",
      "[28]\tInterferon Alfa-2b, Recombinant\t-0.1827530860900879\n",
      "[29]\tMinocycline\t-0.18398095667362213\n",
      "[30]\tPrednisolone\t-0.1842109113931656\n",
      "[31]\tGlutathione\t-0.1865571141242981\n",
      "[32]\tDB14128\t-0.19021636247634888\n",
      "[33]\tFolic Acid\t-0.19326183199882507\n",
      "[34]\tTriamcinolone\t-0.19339141249656677\n",
      "[35]\tDB09341\t-0.1942635327577591\n",
      "[36]\tMycophenolate mofetil\t-0.19588710367679596\n",
      "[37]\tSirolimus\t-0.19725200533866882\n",
      "[38]\tFluoxetine\t-0.19818906486034393\n",
      "[39]\tSulfasalazine\t-0.19829334318637848\n",
      "[40]\tDB09122\t-0.2008952647447586\n",
      "[41]\tDB14104\t-0.20264773070812225\n",
      "[42]\tPazopanib\t-0.20329171419143677\n",
      "[43]\tTestosterone\t-0.20426590740680695\n",
      "[44]\tPeginterferon alfa-2a\t-0.2048606425523758\n",
      "[45]\tDinoprostone\t-0.20563732087612152\n",
      "[46]\tCyclic Adenosine Monophosphate\t-0.2057119607925415\n",
      "[47]\tEculizumab\t-0.20663462579250336\n",
      "[48]\tNatalizumab\t-0.21150510013103485\n",
      "[49]\tHydroxychloroquine\t-0.21465201675891876\n"
     ]
    }
   ],
   "source": [
    "#Hallo las 50 drogas mejor rankeadas junto con sus scores:\n",
    "for i in range(50):\n",
    "    drug_EAD = int(proposed_dids_EAD[i])\n",
    "    drug_code_EAD = entity_id_map[drug_EAD]\n",
    "    DB_code_EAD = drug_code_EAD.split('::')[1]\n",
    "    drugbank_row_EAD = df_drugbank[df_drugbank['drugbank_id'] == DB_code_EAD]['name']\n",
    "    if drugbank_row_EAD.any() == True:\n",
    "      drug_name_EAD = drugbank_row_EAD.item()\n",
    "    else:\n",
    "      drug_name_EAD = DB_code_EAD\n",
    "\n",
    "    score_EAD = proposed_scores_EAD[i]\n",
    "\n",
    "    print(\"[{}]\\t{}\\t{}\".format(i, drug_name_EAD,score_EAD))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "pu-t0WSjTfb0",
   "metadata": {
    "id": "pu-t0WSjTfb0"
   },
   "source": [
    "Hallo en la página de Drugbank los códigos que no figuran en el .tsv:\n",
    "* DB11094 ----> Vitamina D.\n",
    "* DB11672 ----> Curcumina.\n",
    "* DB14783 ----> Diroximel fumarato. (used for the treatmeant relapsing forms of Multiple Sclerosis (MS)).\n",
    "* DB13038 ----> Trioleato de glicerol (Trioleína).\n",
    "* DB14128 ----> Nadide.\n",
    "* DB09341 ----> Glucose.\n",
    "* DB09122 ----> Peginterferon beta-1a.\n",
    "* DB14104 ----> Linoleic acid."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "lEYXPv0MYSqQ",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 479,
     "status": "ok",
     "timestamp": 1729200748992,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "lEYXPv0MYSqQ",
    "outputId": "7fd14925-1ee6-4b76-f784-b8c949bb6f94"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\tMethylprednisolone\t-0.12100826948881149\n",
      "[1]\tAzathioprine\t-0.12112560868263245\n",
      "[2]\tMethotrexate\t-0.12847775220870972\n",
      "[3]\tDexamethasone\t-0.12967625260353088\n",
      "[4]\tMelatonin\t-0.132446750998497\n",
      "[5]\tTeriflunomide\t-0.13346567749977112\n",
      "[6]\tVitamin D\t-0.1414821892976761\n",
      "[7]\tCladribine\t-0.14233370125293732\n",
      "[8]\tThalidomide\t-0.1431959718465805\n",
      "[9]\tEstradiol\t-0.14724911749362946\n",
      "[10]\tCurcumin\t-0.1509384959936142\n",
      "[11]\tDiroximel fumarate\t-0.15210206806659698\n",
      "[12]\tCyclosporine\t-0.15574337542057037\n",
      "[13]\tCholesterol\t-0.15823456645011902\n",
      "[14]\tHydrocortisone\t-0.16551189124584198\n",
      "[15]\tGlyceryl trioleate\t-0.1685316413640976\n",
      "[16]\tMorphine\t-0.1695857048034668\n",
      "[17]\tPrednisone\t-0.16970013082027435\n",
      "[18]\tSimvastatin\t-0.17194552719593048\n",
      "[19]\tIndomethacin\t-0.1736704409122467\n",
      "[20]\tFluticasone Propionate\t-0.17437157034873962\n",
      "[21]\tColchicine\t-0.1745726466178894\n",
      "[22]\tDoxorubicin\t-0.17502200603485107\n",
      "[23]\tMitoxantrone\t-0.1756279468536377\n",
      "[24]\tTretinoin\t-0.1766490787267685\n",
      "[25]\tPentoxifylline\t-0.17967404425144196\n",
      "[26]\tCisplatin\t-0.18140099942684174\n",
      "[27]\tInterferon beta-1b\t-0.1818210333585739\n",
      "[28]\tInterferon Alfa-2b, Recombinant\t-0.1827530860900879\n",
      "[29]\tMinocycline\t-0.18398095667362213\n",
      "[30]\tPrednisolone\t-0.1842109113931656\n",
      "[31]\tGlutathione\t-0.1865571141242981\n",
      "[32]\tNadide\t-0.19021636247634888\n",
      "[33]\tFolic Acid\t-0.19326183199882507\n",
      "[34]\tTriamcinolone\t-0.19339141249656677\n",
      "[35]\tGlucose\t-0.1942635327577591\n",
      "[36]\tMycophenolate mofetil\t-0.19588710367679596\n",
      "[37]\tSirolimus\t-0.19725200533866882\n",
      "[38]\tFluoxetine\t-0.19818906486034393\n",
      "[39]\tSulfasalazine\t-0.19829334318637848\n",
      "[40]\tPeginterferon beta-1a\t-0.2008952647447586\n",
      "[41]\tLinoleic acid\t-0.20264773070812225\n",
      "[42]\tPazopanib\t-0.20329171419143677\n",
      "[43]\tTestosterone\t-0.20426590740680695\n",
      "[44]\tPeginterferon alfa-2a\t-0.2048606425523758\n",
      "[45]\tDinoprostone\t-0.20563732087612152\n",
      "[46]\tCyclic Adenosine Monophosphate\t-0.2057119607925415\n",
      "[47]\tEculizumab\t-0.20663462579250336\n",
      "[48]\tNatalizumab\t-0.21150510013103485\n",
      "[49]\tHydroxychloroquine\t-0.21465201675891876\n"
     ]
    }
   ],
   "source": [
    "for i in range(50):\n",
    "    drug_EAD = int(proposed_dids_EAD[i])\n",
    "    drug_code = entity_id_map[drug_EAD]\n",
    "    DB_code = drug_code.split('::')[1]\n",
    "    drugbank_row = df_drugbank[df_drugbank['drugbank_id'] == DB_code]['name']\n",
    "    if drugbank_row.any() == True:\n",
    "      drug_name = drugbank_row.item()\n",
    "    else:\n",
    "      if DB_code == 'DB11094':\n",
    "        drug_name = 'Vitamin D'\n",
    "      elif DB_code == 'DB11672':\n",
    "        drug_name = 'Curcumin'\n",
    "      elif DB_code == 'DB14783':\n",
    "        drug_name = 'Diroximel fumarate'\n",
    "      elif DB_code == 'DB13038':\n",
    "        drug_name = 'Glyceryl trioleate'\n",
    "      elif DB_code == 'DB14128':\n",
    "        drug_name = 'Nadide'\n",
    "      elif DB_code == 'DB09341':\n",
    "        drug_name = 'Glucose'\n",
    "      elif DB_code == 'DB09122':\n",
    "        drug_name = 'Peginterferon beta-1a'\n",
    "      elif DB_code == 'DB14104':\n",
    "        drug_name = 'Linoleic acid'\n",
    "\n",
    "    score_EAD = proposed_scores_EAD[i]\n",
    "\n",
    "    print(\"[{}]\\t{}\\t{}\".format(i, drug_name,score_EAD))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "EfRr1bEWtgJs",
   "metadata": {
    "id": "EfRr1bEWtgJs"
   },
   "source": [
    "Aplicamos el mismo procedimiento anterior pero solo para los distintos tipos de esclerosis múltiple (MS):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9sWG9afdtiiw",
   "metadata": {
    "id": "9sWG9afdtiiw"
   },
   "outputs": [],
   "source": [
    "lista_MS = [\n",
    "'Disease::MESH:D009103',\n",
    "'Disease::MESH:D020528',\n",
    "'Disease::MESH:D020529'\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Ge1PalrquW1Q",
   "metadata": {
    "id": "Ge1PalrquW1Q"
   },
   "outputs": [],
   "source": [
    "# Sacamos los ID de los tratamientos elegidos del mapa completo de relaciones:\n",
    "disease_ids_MS = []\n",
    "\n",
    "for disease in lista_MS:\n",
    "    disease_ids_MS.append(entity_map[disease])\n",
    "\n",
    "treatment_rid_MS = [relation_map[treat]  for treat in treatment]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9u069OFresFh",
   "metadata": {
    "id": "9u069OFresFh"
   },
   "outputs": [],
   "source": [
    "# Cargamos los embeddings de los tratamientos relacionados a MS:\n",
    "disease_ids_MS = torch.tensor(disease_ids_MS).long()\n",
    "treatment_rid_MS = torch.tensor(treatment_rid_MS)\n",
    "\n",
    "treatment_embs_MS = [torch.tensor(emb_rel[rid]) for rid in treatment_rid_MS]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "RKsfQDSNe8dz",
   "metadata": {
    "id": "RKsfQDSNe8dz"
   },
   "outputs": [],
   "source": [
    "scores_per_disease_MS = []\n",
    "dids_MS = []\n",
    "for rid in range(len(treatment_embs_MS)):\n",
    "    treatment_emb_MS=treatment_embs_MS[rid]\n",
    "    for disease_id in disease_ids_MS:\n",
    "        disease_emb_MS = emb_entity[disease_id]\n",
    "        score_MS = fn.logsigmoid(transE_l2(drug_emb, treatment_emb_MS, disease_emb_MS))\n",
    "        scores_per_disease_MS.append(score_MS)\n",
    "        dids_MS.append(drug_ids)\n",
    "scores_MS = torch.cat(scores_per_disease_MS)\n",
    "dids_MS = torch.cat(dids_MS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "YIaa0msxgHHl",
   "metadata": {
    "id": "YIaa0msxgHHl"
   },
   "outputs": [],
   "source": [
    "# Ordenamos los scores en orden descendente\n",
    "idx_MS = torch.flip(torch.argsort(scores_MS), dims=[0])\n",
    "scores_MS = scores_MS[idx_MS].numpy()\n",
    "dids_MS = dids_MS[idx_MS].numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "uxL8_-spgG5v",
   "metadata": {
    "id": "uxL8_-spgG5v"
   },
   "outputs": [],
   "source": [
    "#Hallamos los tratamientos propuestos:\n",
    "_, unique_indices_MS = np.unique(dids_MS, return_index=True)\n",
    "topk=100\n",
    "topk_indices_MS = np.sort(unique_indices_MS)[:topk]\n",
    "proposed_dids_MS = dids_MS[topk_indices_MS]\n",
    "proposed_scores_MS = scores_MS[topk_indices_MS]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "iOkUbRjvkFm7",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1729200857868,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "iOkUbRjvkFm7",
    "outputId": "6fd0bc34-644f-4dbb-b295-179f9599d42b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\tMethylprednisolone\t-0.12100826948881149\n",
      "[1]\tAzathioprine\t-0.12112560868263245\n",
      "[2]\tMethotrexate\t-0.12847775220870972\n",
      "[3]\tDexamethasone\t-0.12967625260353088\n",
      "[4]\tMelatonin\t-0.132446750998497\n",
      "[5]\tTeriflunomide\t-0.13346567749977112\n",
      "[6]\tDB11094\t-0.1414821892976761\n",
      "[7]\tCladribine\t-0.14233370125293732\n",
      "[8]\tThalidomide\t-0.1431959718465805\n",
      "[9]\tEstradiol\t-0.14724911749362946\n",
      "[10]\tDB11672\t-0.1509384959936142\n",
      "[11]\tDB14783\t-0.15210206806659698\n",
      "[12]\tCyclosporine\t-0.15574337542057037\n",
      "[13]\tCholesterol\t-0.15823456645011902\n",
      "[14]\tHydrocortisone\t-0.16551189124584198\n",
      "[15]\tMorphine\t-0.1695857048034668\n",
      "[16]\tPrednisone\t-0.16970013082027435\n",
      "[17]\tSimvastatin\t-0.17194552719593048\n",
      "[18]\tIndomethacin\t-0.1736704409122467\n",
      "[19]\tFluticasone Propionate\t-0.17437157034873962\n",
      "[20]\tColchicine\t-0.1745726466178894\n",
      "[21]\tDoxorubicin\t-0.17502200603485107\n",
      "[22]\tMitoxantrone\t-0.1756279468536377\n",
      "[23]\tTretinoin\t-0.1766490787267685\n",
      "[24]\tPentoxifylline\t-0.17967404425144196\n",
      "[25]\tCisplatin\t-0.18140099942684174\n",
      "[26]\tInterferon beta-1b\t-0.1818210333585739\n",
      "[27]\tInterferon Alfa-2b, Recombinant\t-0.1827530860900879\n",
      "[28]\tMinocycline\t-0.18398095667362213\n",
      "[29]\tPrednisolone\t-0.1842109113931656\n",
      "[30]\tGlutathione\t-0.1865571141242981\n",
      "[31]\tDB14128\t-0.19021636247634888\n",
      "[32]\tFolic Acid\t-0.19326183199882507\n",
      "[33]\tTriamcinolone\t-0.19339141249656677\n",
      "[34]\tDB09341\t-0.1942635327577591\n",
      "[35]\tMycophenolate mofetil\t-0.19588710367679596\n",
      "[36]\tSirolimus\t-0.19725200533866882\n",
      "[37]\tFluoxetine\t-0.19818906486034393\n",
      "[38]\tSulfasalazine\t-0.19829334318637848\n",
      "[39]\tDB09122\t-0.2008952647447586\n",
      "[40]\tDB14104\t-0.20264773070812225\n",
      "[41]\tPazopanib\t-0.20329171419143677\n",
      "[42]\tTestosterone\t-0.20426590740680695\n",
      "[43]\tPeginterferon alfa-2a\t-0.2048606425523758\n",
      "[44]\tDinoprostone\t-0.20563732087612152\n",
      "[45]\tCyclic Adenosine Monophosphate\t-0.2057119607925415\n",
      "[46]\tNatalizumab\t-0.21150510013103485\n",
      "[47]\tHydroxychloroquine\t-0.21465201675891876\n",
      "[48]\tEstropipate\t-0.21523913741111755\n",
      "[49]\tTopotecan\t-0.21528679132461548\n"
     ]
    }
   ],
   "source": [
    "for i in range(50):\n",
    "    drug_MS = int(proposed_dids_MS[i])\n",
    "    drug_code_MS = entity_id_map[drug_MS]\n",
    "    DB_code_MS = drug_code_MS.split('::')[1]\n",
    "    drugbank_row_MS = df_drugbank[df_drugbank['drugbank_id'] == DB_code_MS]['name']\n",
    "    if drugbank_row_MS.any() == True:\n",
    "      drug_name_MS = drugbank_row_MS.item()\n",
    "    else:\n",
    "      drug_name_MS = DB_code_MS\n",
    "\n",
    "    score_ms = proposed_scores_MS[i]\n",
    "\n",
    "    print(\"[{}]\\t{}\\t{}\".format(i, drug_name_MS,score_ms))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6MrHft7DZ5mt",
   "metadata": {
    "id": "6MrHft7DZ5mt"
   },
   "source": [
    "Hallo en la página de Drugbank los códigos que no figuran en el .tsv e imprimimos los resultados con los nombres faltantes:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "-qOD1k9NmMZ1",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 472,
     "status": "ok",
     "timestamp": 1729200918722,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "-qOD1k9NmMZ1",
    "outputId": "5eb4be33-33a7-4df2-9134-a2227b159b95"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\tMethylprednisolone\t-0.12100826948881149\n",
      "[1]\tAzathioprine\t-0.12112560868263245\n",
      "[2]\tMethotrexate\t-0.12847775220870972\n",
      "[3]\tDexamethasone\t-0.12967625260353088\n",
      "[4]\tMelatonin\t-0.132446750998497\n",
      "[5]\tTeriflunomide\t-0.13346567749977112\n",
      "[6]\tVitamin D\t-0.1414821892976761\n",
      "[7]\tCladribine\t-0.14233370125293732\n",
      "[8]\tThalidomide\t-0.1431959718465805\n",
      "[9]\tEstradiol\t-0.14724911749362946\n",
      "[10]\tCurcumin\t-0.1509384959936142\n",
      "[11]\tDiroximel fumarate\t-0.15210206806659698\n",
      "[12]\tCyclosporine\t-0.15574337542057037\n",
      "[13]\tCholesterol\t-0.15823456645011902\n",
      "[14]\tHydrocortisone\t-0.16551189124584198\n",
      "[15]\tMorphine\t-0.1695857048034668\n",
      "[16]\tPrednisone\t-0.16970013082027435\n",
      "[17]\tSimvastatin\t-0.17194552719593048\n",
      "[18]\tIndomethacin\t-0.1736704409122467\n",
      "[19]\tFluticasone Propionate\t-0.17437157034873962\n",
      "[20]\tColchicine\t-0.1745726466178894\n",
      "[21]\tDoxorubicin\t-0.17502200603485107\n",
      "[22]\tMitoxantrone\t-0.1756279468536377\n",
      "[23]\tTretinoin\t-0.1766490787267685\n",
      "[24]\tPentoxifylline\t-0.17967404425144196\n",
      "[25]\tCisplatin\t-0.18140099942684174\n",
      "[26]\tInterferon beta-1b\t-0.1818210333585739\n",
      "[27]\tInterferon Alfa-2b, Recombinant\t-0.1827530860900879\n",
      "[28]\tMinocycline\t-0.18398095667362213\n",
      "[29]\tPrednisolone\t-0.1842109113931656\n",
      "[30]\tGlutathione\t-0.1865571141242981\n",
      "[31]\tNadide\t-0.19021636247634888\n",
      "[32]\tFolic Acid\t-0.19326183199882507\n",
      "[33]\tTriamcinolone\t-0.19339141249656677\n",
      "[34]\tGlucose\t-0.1942635327577591\n",
      "[35]\tMycophenolate mofetil\t-0.19588710367679596\n",
      "[36]\tSirolimus\t-0.19725200533866882\n",
      "[37]\tFluoxetine\t-0.19818906486034393\n",
      "[38]\tSulfasalazine\t-0.19829334318637848\n",
      "[39]\tPeginterferon beta-1a\t-0.2008952647447586\n",
      "[40]\tLinoleic Acid\t-0.20264773070812225\n",
      "[41]\tPazopanib\t-0.20329171419143677\n",
      "[42]\tTestosterone\t-0.20426590740680695\n",
      "[43]\tPeginterferon alfa-2a\t-0.2048606425523758\n",
      "[44]\tDinoprostone\t-0.20563732087612152\n",
      "[45]\tCyclic Adenosine Monophosphate\t-0.2057119607925415\n",
      "[46]\tNatalizumab\t-0.21150510013103485\n",
      "[47]\tHydroxychloroquine\t-0.21465201675891876\n",
      "[48]\tEstropipate\t-0.21523913741111755\n",
      "[49]\tTopotecan\t-0.21528679132461548\n"
     ]
    }
   ],
   "source": [
    "for i in range(50):\n",
    "    drug_MS = int(proposed_dids_MS[i])\n",
    "    drug_code_MS = entity_id_map[drug_MS]\n",
    "    DB_code_MS = drug_code_MS.split('::')[1]\n",
    "    drugbank_row_MS = df_drugbank[df_drugbank['drugbank_id'] == DB_code_MS]['name']\n",
    "    if drugbank_row_MS.any() == True:\n",
    "      drug_name_MS = drugbank_row_MS.item()\n",
    "    else:\n",
    "      if DB_code_MS == 'DB11094':\n",
    "        drug_name_MS = 'Vitamin D'\n",
    "      elif DB_code_MS == 'DB11672':\n",
    "        drug_name_MS = 'Curcumin'\n",
    "      elif DB_code_MS == 'DB14783':\n",
    "        drug_name_MS = 'Diroximel fumarate'\n",
    "      elif DB_code_MS == 'DB14128':\n",
    "        drug_name_MS = 'Nadide'\n",
    "      elif DB_code_MS == 'DB09341':\n",
    "        drug_name_MS = 'Glucose'\n",
    "      elif DB_code_MS == 'DB09122':\n",
    "        drug_name_MS = 'Peginterferon beta-1a'\n",
    "      elif DB_code_MS == 'DB14104':\n",
    "        drug_name_MS = 'Linoleic Acid'\n",
    "    score_ms = proposed_scores_MS[i]\n",
    "\n",
    "    print(\"[{}]\\t{}\\t{}\".format(i, drug_name_MS,score_ms))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21hAU00LaJR5",
   "metadata": {
    "id": "21hAU00LaJR5"
   },
   "source": [
    "Idem pero solo para las enfermedades autoinmunes desmielinizantes no esclerosis múltiple (no MS):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "hYRTOEVqhMlo",
   "metadata": {
    "id": "hYRTOEVqhMlo"
   },
   "outputs": [],
   "source": [
    "lista_no_ms = [\n",
    "'Disease::MESH:D002549',\n",
    "'Disease::MESH:D009188',\n",
    "'Disease::MESH:D009471'\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "_hA80AkMhhBd",
   "metadata": {
    "id": "_hA80AkMhhBd"
   },
   "outputs": [],
   "source": [
    "# Sacamos los ID de los tratamientos elegidos del mapa completo de relaciones:\n",
    "disease_ids_no_ms = []\n",
    "\n",
    "for disease in lista_no_ms:\n",
    "    disease_ids_no_ms.append(entity_map[disease])\n",
    "\n",
    "treatment_rid_no_ms = [relation_map[treat]  for treat in treatment]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "AJyOsIWwhu6D",
   "metadata": {
    "id": "AJyOsIWwhu6D"
   },
   "outputs": [],
   "source": [
    "# Cargamos los embeddings de los tratamientos relacionados a MS:\n",
    "disease_ids_no_ms = torch.tensor(disease_ids_no_ms).long()\n",
    "treatment_rid_no_ms = torch.tensor(treatment_rid_no_ms)\n",
    "\n",
    "treatment_embs_no_ms = [torch.tensor(emb_rel[rid]) for rid in treatment_rid_no_ms]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "85DzKoCihuxO",
   "metadata": {
    "id": "85DzKoCihuxO"
   },
   "outputs": [],
   "source": [
    "scores_per_disease_no_ms = []\n",
    "dids_no_ms = []\n",
    "for rid in range(len(treatment_embs_no_ms)):\n",
    "    treatment_emb_no_ms=treatment_embs_no_ms[rid]\n",
    "    for disease_id in disease_ids_no_ms:\n",
    "        disease_emb_no_ms = emb_entity[disease_id]\n",
    "        score_no_ms = fn.logsigmoid(transE_l2(drug_emb, treatment_emb_no_ms, disease_emb_no_ms))\n",
    "        scores_per_disease_no_ms.append(score_no_ms)\n",
    "        dids_no_ms.append(drug_ids)\n",
    "scores_no_ms = torch.cat(scores_per_disease_no_ms)\n",
    "dids_no_ms = torch.cat(dids_no_ms)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "JglK7c_uimVf",
   "metadata": {
    "id": "JglK7c_uimVf"
   },
   "outputs": [],
   "source": [
    "# Ordenamos los scores en orden descendente\n",
    "idx_no_ms = torch.flip(torch.argsort(scores_no_ms), dims=[0])\n",
    "scores_no_ms = scores_no_ms[idx_no_ms].numpy()\n",
    "dids_no_ms = dids_no_ms[idx_no_ms].numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "_B2ckIoUimM0",
   "metadata": {
    "id": "_B2ckIoUimM0"
   },
   "outputs": [],
   "source": [
    "#Hallamos los tratamientos propuestos:\n",
    "_, unique_indices_no_ms = np.unique(dids_no_ms, return_index=True)\n",
    "topk=100\n",
    "topk_indices_no_ms = np.sort(unique_indices_no_ms)[:topk]\n",
    "proposed_dids_no_ms = dids_no_ms[topk_indices_no_ms]\n",
    "proposed_scores_no_ms = scores_no_ms[topk_indices_no_ms]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "FNIRfEjDjOSz",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 561,
     "status": "ok",
     "timestamp": 1729201052226,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "FNIRfEjDjOSz",
    "outputId": "51411182-e80b-46dd-8b56-46b1bb31498d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\tAzathioprine\t-0.16528524458408356\n",
      "[1]\tDB13038\t-0.1685316413640976\n",
      "[2]\tMethylprednisolone\t-0.18176332116127014\n",
      "[3]\tEculizumab\t-0.20663462579250336\n",
      "[4]\tPrednisolone\t-0.20898757874965668\n",
      "[5]\tDexamethasone\t-0.2263961285352707\n",
      "[6]\tMethotrexate\t-0.2489393800497055\n",
      "[7]\tMycophenolate mofetil\t-0.26964807510375977\n",
      "[8]\tCortisone acetate\t-0.312658429145813\n",
      "[9]\tColchicine\t-0.316200852394104\n",
      "[10]\tAmiodarone\t-0.3248213529586792\n",
      "[11]\tPrednisone\t-0.326887309551239\n",
      "[12]\tMitoxantrone\t-0.32996848225593567\n",
      "[13]\tTriamcinolone\t-0.3480216860771179\n",
      "[14]\tMycophenolic acid\t-0.3493836522102356\n",
      "[15]\tPentoxifylline\t-0.3496331572532654\n",
      "[16]\tBudesonide\t-0.35124772787094116\n",
      "[17]\tSirolimus\t-0.36869800090789795\n",
      "[18]\tCorticotropin\t-0.368741512298584\n",
      "[19]\tCholesterol\t-0.3702375292778015\n",
      "[20]\tOfloxacin\t-0.371698260307312\n",
      "[21]\tFludarabine\t-0.3886129558086395\n",
      "[22]\tHydrocortisone\t-0.39652618765830994\n",
      "[23]\tDactinomycin\t-0.3971503674983978\n",
      "[24]\tDoxorubicin\t-0.4012955129146576\n",
      "[25]\tHydroxychloroquine\t-0.41137802600860596\n",
      "[26]\tDaunorubicin\t-0.4163414537906647\n",
      "[27]\tEstradiol\t-0.43716147541999817\n",
      "[28]\tMetoclopramide\t-0.44733795523643494\n",
      "[29]\tAldosterone\t-0.4480136036872864\n",
      "[30]\tIrinotecan\t-0.4487730860710144\n",
      "[31]\tClarithromycin\t-0.4493292272090912\n",
      "[32]\tThalidomide\t-0.4501630961894989\n",
      "[33]\tPaclitaxel\t-0.45132994651794434\n",
      "[34]\tLeflunomide\t-0.4514501690864563\n",
      "[35]\tCyclosporine\t-0.46148377656936646\n",
      "[36]\tMelphalan\t-0.4621807932853699\n",
      "[37]\tABT-874\t-0.46643462777137756\n",
      "[38]\tAminophylline\t-0.46833670139312744\n",
      "[39]\tTopotecan\t-0.4696570336818695\n",
      "[40]\tBetamethasone\t-0.46973493695259094\n",
      "[41]\tTretinoin\t-0.4747369587421417\n",
      "[42]\tSulfasalazine\t-0.47544634342193604\n",
      "[43]\tEtoposide\t-0.4754723310470581\n",
      "[44]\tPazopanib\t-0.48153433203697205\n",
      "[45]\tTacrolimus\t-0.4816385805606842\n",
      "[46]\tAdenosine\t-0.4870721697807312\n",
      "[47]\tMometasone\t-0.4902561604976654\n",
      "[48]\tFenofibrate\t-0.4993910789489746\n",
      "[49]\tVincristine\t-0.5033408999443054\n",
      "[50]\tRifampicin\t-0.5054424405097961\n",
      "[51]\tEstropipate\t-0.5079152584075928\n",
      "[52]\tFluocinolone Acetonide\t-0.5086798667907715\n",
      "[53]\tGlatiramer Acetate\t-0.5125245451927185\n",
      "[54]\tFluticasone Propionate\t-0.5134993195533752\n",
      "[55]\tDasatinib\t-0.5186044573783875\n",
      "[56]\tLevofloxacin\t-0.5231633186340332\n",
      "[57]\tEpirubicin\t-0.5241122245788574\n",
      "[58]\tDocetaxel\t-0.5278612375259399\n",
      "[59]\tDipyridamole\t-0.5315543413162231\n",
      "[60]\tClomifene\t-0.532511293888092\n",
      "[61]\tSimvastatin\t-0.5340142250061035\n",
      "[62]\tPemetrexed\t-0.5347414612770081\n",
      "[63]\tDB06774\t-0.5359562635421753\n",
      "[64]\tTrilostane\t-0.5361312031745911\n",
      "[65]\tFludrocortisone\t-0.5424282550811768\n",
      "[66]\tKetoconazole\t-0.5444682836532593\n",
      "[67]\tEthinyl Estradiol\t-0.5498777031898499\n",
      "[68]\tCiclesonide\t-0.5563293695449829\n",
      "[69]\tGemcitabine\t-0.5614726543426514\n",
      "[70]\tNaloxone\t-0.5622051954269409\n",
      "[71]\tOndansetron\t-0.5629396438598633\n",
      "[72]\tRibavirin\t-0.5652408599853516\n",
      "[73]\tRoflumilast\t-0.5659238696098328\n",
      "[74]\tVinblastine\t-0.5663394331932068\n",
      "[75]\tAfelimomab\t-0.5675961375236511\n",
      "[76]\tCarboplatin\t-0.5690613985061646\n",
      "[77]\tDisulfiram\t-0.5718541145324707\n",
      "[78]\tParamethasone\t-0.5730870366096497\n",
      "[79]\tPentostatin\t-0.5752649903297424\n",
      "[80]\tSunitinib\t-0.5785483717918396\n",
      "[81]\tQuinidine\t-0.5791858434677124\n",
      "[82]\tDB09033\t-0.5792458057403564\n",
      "[83]\tTocilizumab\t-0.5804287195205688\n",
      "[84]\tTeriflunomide\t-0.5834261178970337\n",
      "[85]\tIsotretinoin\t-0.5879727602005005\n",
      "[86]\tAbatacept\t-0.5891955494880676\n",
      "[87]\tOuabain\t-0.5918434262275696\n",
      "[88]\tSparfloxacin\t-0.5929718613624573\n",
      "[89]\tIdarubicin\t-0.5932427644729614\n",
      "[90]\tGenistein\t-0.5947027206420898\n",
      "[91]\tZidovudine\t-0.5999560952186584\n",
      "[92]\tQuinacrine\t-0.6021620035171509\n",
      "[93]\tGemfibrozil\t-0.6023479700088501\n",
      "[94]\tHydrochlorothiazide\t-0.6080540418624878\n",
      "[95]\tItraconazole\t-0.6103712916374207\n",
      "[96]\tRitodrine\t-0.6108227968215942\n",
      "[97]\tNilotinib\t-0.6112086772918701\n",
      "[98]\tSulfamethoxazole\t-0.6116475462913513\n",
      "[99]\tCilostazol\t-0.6117077469825745\n"
     ]
    }
   ],
   "source": [
    "for i in range(topk):\n",
    "    drug_no_ms = int(proposed_dids_no_ms[i])\n",
    "    drug_code_no_ms = entity_id_map[drug_no_ms]\n",
    "    DB_code_no_ms = drug_code_no_ms.split('::')[1]\n",
    "    drugbank_row_no_ms = df_drugbank[df_drugbank['drugbank_id'] == DB_code_no_ms]['name']\n",
    "    if drugbank_row_no_ms.any() == True:\n",
    "      drug_name_no_ms = drugbank_row_no_ms.item()\n",
    "    else:\n",
    "      drug_name_no_ms = DB_code_no_ms\n",
    "\n",
    "    score_no_ms = proposed_scores_no_ms[i]\n",
    "\n",
    "    print(\"[{}]\\t{}\\t{}\".format(i, drug_name_no_ms,score_no_ms))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1vC2htK6asvo",
   "metadata": {
    "id": "1vC2htK6asvo"
   },
   "source": [
    "Hallo en la página de Drugbank los códigos que no figuran en el .tsv e imprimimos los resultados con los nombres faltantes:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "T_cwxYjsjOGB",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 464,
     "status": "ok",
     "timestamp": 1729201103498,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "T_cwxYjsjOGB",
    "outputId": "468714a8-244e-4788-f392-d5180951153a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\tAzathioprine\t-0.16528524458408356\n",
      "[1]\tGlyceryl Trioleate\t-0.1685316413640976\n",
      "[2]\tMethylprednisolone\t-0.18176332116127014\n",
      "[3]\tEculizumab\t-0.20663462579250336\n",
      "[4]\tPrednisolone\t-0.20898757874965668\n",
      "[5]\tDexamethasone\t-0.2263961285352707\n",
      "[6]\tMethotrexate\t-0.2489393800497055\n",
      "[7]\tMycophenolate mofetil\t-0.26964807510375977\n",
      "[8]\tCortisone acetate\t-0.312658429145813\n",
      "[9]\tColchicine\t-0.316200852394104\n",
      "[10]\tAmiodarone\t-0.3248213529586792\n",
      "[11]\tPrednisone\t-0.326887309551239\n",
      "[12]\tMitoxantrone\t-0.32996848225593567\n",
      "[13]\tTriamcinolone\t-0.3480216860771179\n",
      "[14]\tMycophenolic acid\t-0.3493836522102356\n",
      "[15]\tPentoxifylline\t-0.3496331572532654\n",
      "[16]\tBudesonide\t-0.35124772787094116\n",
      "[17]\tSirolimus\t-0.36869800090789795\n",
      "[18]\tCorticotropin\t-0.368741512298584\n",
      "[19]\tCholesterol\t-0.3702375292778015\n",
      "[20]\tOfloxacin\t-0.371698260307312\n",
      "[21]\tFludarabine\t-0.3886129558086395\n",
      "[22]\tHydrocortisone\t-0.39652618765830994\n",
      "[23]\tDactinomycin\t-0.3971503674983978\n",
      "[24]\tDoxorubicin\t-0.4012955129146576\n",
      "[25]\tHydroxychloroquine\t-0.41137802600860596\n",
      "[26]\tDaunorubicin\t-0.4163414537906647\n",
      "[27]\tEstradiol\t-0.43716147541999817\n",
      "[28]\tMetoclopramide\t-0.44733795523643494\n",
      "[29]\tAldosterone\t-0.4480136036872864\n",
      "[30]\tIrinotecan\t-0.4487730860710144\n",
      "[31]\tClarithromycin\t-0.4493292272090912\n",
      "[32]\tThalidomide\t-0.4501630961894989\n",
      "[33]\tPaclitaxel\t-0.45132994651794434\n",
      "[34]\tLeflunomide\t-0.4514501690864563\n",
      "[35]\tCyclosporine\t-0.46148377656936646\n",
      "[36]\tMelphalan\t-0.4621807932853699\n",
      "[37]\tABT-874\t-0.46643462777137756\n",
      "[38]\tAminophylline\t-0.46833670139312744\n",
      "[39]\tTopotecan\t-0.4696570336818695\n",
      "[40]\tBetamethasone\t-0.46973493695259094\n",
      "[41]\tTretinoin\t-0.4747369587421417\n",
      "[42]\tSulfasalazine\t-0.47544634342193604\n",
      "[43]\tEtoposide\t-0.4754723310470581\n",
      "[44]\tPazopanib\t-0.48153433203697205\n",
      "[45]\tTacrolimus\t-0.4816385805606842\n",
      "[46]\tAdenosine\t-0.4870721697807312\n",
      "[47]\tMometasone\t-0.4902561604976654\n",
      "[48]\tFenofibrate\t-0.4993910789489746\n",
      "[49]\tVincristine\t-0.5033408999443054\n"
     ]
    }
   ],
   "source": [
    "for i in range(50):\n",
    "    drug_no_ms = int(proposed_dids_no_ms[i])\n",
    "    drug_code_no_ms = entity_id_map[drug_no_ms]\n",
    "    DB_code_no_ms = drug_code_no_ms.split('::')[1]\n",
    "    drugbank_row_no_ms = df_drugbank[df_drugbank['drugbank_id'] == DB_code_no_ms]['name']\n",
    "    if drugbank_row_no_ms.any() == True:\n",
    "      drug_name_no_ms = drugbank_row_no_ms.item()\n",
    "    else:\n",
    "      if DB_code_no_ms == 'DB13038':\n",
    "        drug_name_no_ms = 'Glyceryl Trioleate'\n",
    "\n",
    "    score_no_ms = proposed_scores_no_ms[i]\n",
    "\n",
    "    print(\"[{}]\\t{}\\t{}\".format(i, drug_name_no_ms,score_no_ms))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "IPhgEWzIrjP5",
   "metadata": {
    "id": "IPhgEWzIrjP5"
   },
   "source": [
    "Generamos un archivo .tsv en base a las 100 drogas mejor rankeadas de cada variante analizada (repurposing para las EADs en general, solo para esclerosis múltiple y solo para las EADs no esclerosis múltiple) a fin de tener todas estas drogas propuestas por el algoritmo como base para hacer el mismo análisis pero tomando en cuenta las relaciones gen-enfermedad/enfermedad-gen."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8U6CTjFX7PCk",
   "metadata": {
    "id": "8U6CTjFX7PCk"
   },
   "outputs": [],
   "source": [
    "proposed_dids_aux = np.concatenate((proposed_dids_EAD,proposed_dids_MS,proposed_dids_no_ms)) #Construyo np.array uniendo las 100 drogas propuestas para cada grupo de enfermedades"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "UJ38noF48D2k",
   "metadata": {
    "id": "UJ38noF48D2k"
   },
   "outputs": [],
   "source": [
    "proposed_dids_tot = np.unique(proposed_dids_aux) #Me quedo solo con los elementos únicos del arreglo anterior para no repetir drogas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2iw_T9q-8oaT",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1025,
     "status": "ok",
     "timestamp": 1729201157072,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "2iw_T9q-8oaT",
    "outputId": "99eccc8d-f737-48e3-98cf-a81c467a7c9a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(300,) (154,)\n"
     ]
    }
   ],
   "source": [
    "print(np.shape(proposed_dids_aux),np.shape(proposed_dids_tot)) #Verifico tamaños de los arrays anteriores. Hay 154 drogas diferentes del conjunto inicial de 300"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "WxDvffqwrhco",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1500,
     "status": "ok",
     "timestamp": 1729201191418,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "WxDvffqwrhco",
    "outputId": "bb029591-4161-48b5-87ea-9ad6c72be1e9"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Archivo drogas_repurposing.tsv creado\n"
     ]
    }
   ],
   "source": [
    "with open(\"drogas_repurposing.tsv\", mode=\"w\", newline='') as file: \n",
    "    tsv_writer = csv.writer(file, delimiter='\\t')\n",
    "\n",
    "    # Defino el nombre de las columnas:\n",
    "    tsv_writer.writerow([\"Index\", \"Drug Name\", \"ID Drugbank\"])\n",
    "\n",
    "    for i in range(len(proposed_dids_tot)):\n",
    "        drug_code_rep = entity_id_map[proposed_dids_tot[i]]\n",
    "        DB_code_rep = drug_code_rep.split('::')[1]\n",
    "        drugbank_row_rep = df_drugbank[df_drugbank['drugbank_id'] == DB_code_rep]['name']\n",
    "\n",
    "        if drugbank_row_rep.any() == True:\n",
    "            drug_name = drugbank_row_rep.item()\n",
    "        elif DB_code_rep == 'DB11094':\n",
    "            drug_name = 'Vitamin D'\n",
    "        elif DB_code_rep == 'DB11672':\n",
    "            drug_name = 'Curcumin'\n",
    "        elif DB_code_rep == 'DB14783':\n",
    "            drug_name = 'Diroximel fumarate'\n",
    "        elif DB_code_rep == 'DB13038':\n",
    "            drug_name = 'Glyceryl trioleate'\n",
    "        elif DB_code_rep == 'DB14128':\n",
    "            drug_name = 'Nadide'\n",
    "        elif DB_code_rep == 'DB09341':\n",
    "            drug_name = 'Glucose'\n",
    "        elif DB_code_rep == 'DB09122':\n",
    "            drug_name = 'Peginterferon beta-1a'\n",
    "        elif DB_code_rep == 'DB14104':\n",
    "            drug_name = 'Linoleic acid'\n",
    "        else:\n",
    "            drug_name = DB_code_rep #En estos casos, cuando se haga el análisis posterior con relaciones gen-droga/droga-gen y alguna de las drogas ppales tenga como nombre su ID\n",
    "                                    #de DrugBank, se buscará en la página de dicha base de datos el nombre de compuesto.\n",
    "\n",
    "        index = i + 1\n",
    "\n",
    "        # Escribimos la información en el archivo .tsv\n",
    "        tsv_writer.writerow([index, drug_name, DB_code_rep])\n",
    "\n",
    "print(\"Archivo drogas_repurposing.tsv creado\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "MLgOEg8drN9j",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 486,
     "status": "ok",
     "timestamp": 1729201842207,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "MLgOEg8drN9j",
    "outputId": "c91292b2-7267-4bd7-cd1a-9103509981cc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Losartan:\n",
      "            Entity_name  Entity_ID\n",
      "7897  Compound::DB00678       7897\n",
      "Atorvastatin:\n",
      "            Entity_name  Entity_ID\n",
      "6030  Compound::DB01076       6030\n",
      "Filgrastim:\n",
      "             Entity_name  Entity_ID\n",
      "10587  Compound::DB00099      10587\n",
      "Gemcitabine:\n",
      "            Entity_name  Entity_ID\n",
      "9743  Compound::DB00441       9743\n",
      "Azithromycin:\n",
      "            Entity_name  Entity_ID\n",
      "8552  Compound::DB00207       8552\n",
      "Trichloroacetic acid:\n",
      "                  Entity_name  Entity_ID\n",
      "49605  Compound::MESH:D014238      49605\n",
      "Advanced glycation end-products(AGEs):\n",
      "Empty DataFrame\n",
      "Columns: [Entity_name, Entity_ID]\n",
      "Index: []\n",
      "Amyloid ß:\n",
      "Empty DataFrame\n",
      "Columns: [Entity_name, Entity_ID]\n",
      "Index: []\n"
     ]
    }
   ],
   "source": [
    "#Las drogas propuestas para repurposing por el algoritmo RDKG-115 (https://github.com/ZhuChaoY/RDKG-115) que no figuran entre las drogas propuestas por el algoritmo de DRKG para las enfermedades analizadas al aplicar el modelo,\n",
    "#están en DRKG o no??\n",
    "#Ejemplo de presentación de una droga en el DRKG:\tCompound::DB13410, Compound::MESH:C106125, Compound::CHEMBL481806.\n",
    "\n",
    "#losartan: DB00678, D019808, CHEMBL191.\n",
    "print('Losartan:')\n",
    "print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('DB00678')]) #Aparece solo con este código de DrugBank\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('D019808')])\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('CHEMBL191')])\n",
    "\n",
    "#Atorvastatin: DB01076, D000069059, CHEMBL1487.\n",
    "print('Atorvastatin:')\n",
    "print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('DB01076')]) #Aparece solo con este código de DrugBank\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('D019808')])\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('CHEMBL191')])\n",
    "\n",
    "#Filgrastim: DB00099, D000069585, CHEMBL1201567.\n",
    "print('Filgrastim:')\n",
    "print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('DB00099')]) #Aparece solo con este código de DrugBank\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('D000069585')])\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('CHEMBL1201567')])\n",
    "\n",
    "#Gemcitabine: DB00441, D000093542, CHEMBL888.\n",
    "print('Gemcitabine:')\n",
    "print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('DB00441')]) #Aparece solo con este código de DrugBank\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('D000093542')])\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('CHEMBL888')])\n",
    "\n",
    "#Azithromycin: DB00207, D017963, CHEMBL529.\n",
    "print('Azithromycin:')\n",
    "print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('DB00207')]) #Aparece solo con este código de DrugBank\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('D017963')])\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('CHEMBL529')])\n",
    "\n",
    "#Trichloroacetic acid: DB11152, D014238, CHEMBL14053.\n",
    "print('Trichloroacetic acid:')\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('DB11152')]) #Aparece solo con este código de DrugBank\n",
    "print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('D014238')])\n",
    "#print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('CHEMBL14053')])\n",
    "\n",
    "#Advanced glycation end-products(AGEs): D017127\n",
    "print('Advanced glycation end-products(AGEs):')\n",
    "print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('D017127')])\n",
    "\n",
    "#Amyloid ß: D016229\n",
    "print('Amyloid ß:')\n",
    "print(entidades_drugs[entidades_drugs['Entity_name'].str.endswith('D016229')])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "isUUAO7qbvz4",
   "metadata": {
    "id": "isUUAO7qbvz4"
   },
   "source": [
    "Todas salvo 2 de las drogas de las propuestas por el algoritmo alternativo RDKG-115 (no presente en este código) no están presentes en nuestro modelo. Sin embargo las restantes no son propuestas entre las primeras opciones para repurposing por el DRKG!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ddkAZu2reKpK",
   "metadata": {
    "id": "ddkAZu2reKpK"
   },
   "source": [
    "#Drug Repurposing aplicado a las enfermedades autoinmunes desmielinizantes a través de relaciones entre genes y drogas."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "h8bVc9TBVnAs",
   "metadata": {
    "id": "h8bVc9TBVnAs"
   },
   "source": [
    "Los genes se identifican a través de IDs denominados Entrez Gene ID.\n",
    "Entrez Gene genera números enteros únicos (GeneID) como identificadores estables para genes y otros Locus (posición en el cromosoma donde está localizado un gen específico) para un subconjunto de organismos modelo. Los identificadores se usan para integrar múltiples tipos de información incluyendo nomenclatura, descripciones resumidas, etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "_tQ_VkLwL_bD",
   "metadata": {
    "id": "_tQ_VkLwL_bD"
   },
   "outputs": [],
   "source": [
    "df_gen_eads = pd.read_excel('./EDA_GENES/GENES EAD.xlsx') #GENES EAD.xlsx contiene un listado de los genes relacionados con las EADs obtenido de la página https://www.malacards.org/\n",
    "                                                          #Especificar la carpeta donde está guardado el archivo!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "qJTamfn-L_UJ",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 470,
     "status": "ok",
     "timestamp": 1729202217053,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "qJTamfn-L_UJ",
    "outputId": "756637f5-4ba2-452f-9d75-b2a7ebecd901"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "summary": "{\n  \"name\": \"df_gen_eads\",\n  \"rows\": 191,\n  \"fields\": [\n    {\n      \"column\": \"#\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 55,\n        \"min\": 1,\n        \"max\": 191,\n        \"num_unique_values\": 191,\n        \"samples\": [\n          153,\n          76,\n          159\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Symbol\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 190,\n        \"samples\": [\n          \"NPHP1\",\n          \"RPGRIP1L\",\n          \"CCL3\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Description\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 191,\n        \"samples\": [\n          \"Nectin Cell Adhesion Molecule 2\",\n          \"C-X-C Motif Chemokine Ligand 10\",\n          \"BCL6 Corepressor Like 1\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"ID GEN\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 20243185,\n        \"min\": 25,\n        \"max\": 104326192,\n        \"num_unique_values\": 190,\n        \"samples\": [\n          4867,\n          23322,\n          6348\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
       "type": "dataframe",
       "variable_name": "df_gen_eads"
      },
      "text/html": [
       "\n",
       "  <div id=\"df-8ceff38a-9da5-45b0-a2a1-2fbb948e6c94\" class=\"colab-df-container\">\n",
       "    <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>#</th>\n",
       "      <th>Symbol</th>\n",
       "      <th>Description</th>\n",
       "      <th>ID GEN</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>ABCD1</td>\n",
       "      <td>ATP Binding Cassette Subfamily D Member 1</td>\n",
       "      <td>215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>GALC</td>\n",
       "      <td>Galactosylceramidase</td>\n",
       "      <td>2581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>CSF1</td>\n",
       "      <td>Colony Stimulating Factor 1</td>\n",
       "      <td>1435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>NTF3</td>\n",
       "      <td>Neurotrophin 3</td>\n",
       "      <td>4908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>MIR92A1</td>\n",
       "      <td>MicroRNA 92a-1</td>\n",
       "      <td>407048</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8ceff38a-9da5-45b0-a2a1-2fbb948e6c94')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-8ceff38a-9da5-45b0-a2a1-2fbb948e6c94 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-8ceff38a-9da5-45b0-a2a1-2fbb948e6c94');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-dbce87a7-d215-49eb-b369-413789ea77e0\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-dbce87a7-d215-49eb-b369-413789ea77e0')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-dbce87a7-d215-49eb-b369-413789ea77e0 button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "   #   Symbol                                Description  ID GEN\n",
       "0  1    ABCD1  ATP Binding Cassette Subfamily D Member 1     215\n",
       "1  2     GALC                       Galactosylceramidase    2581\n",
       "2  3     CSF1                Colony Stimulating Factor 1    1435\n",
       "3  4     NTF3                             Neurotrophin 3    4908\n",
       "4  5  MIR92A1                             MicroRNA 92a-1  407048"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_gen_eads.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "tEgr3QNHL_LS",
   "metadata": {
    "collapsed": true,
    "id": "tEgr3QNHL_LS"
   },
   "outputs": [],
   "source": [
    "lista_id_genes = (df_gen_eads['ID GEN']).tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "71UYhZLpcmiA",
   "metadata": {
    "id": "71UYhZLpcmiA"
   },
   "outputs": [],
   "source": [
    "lista_genes_EAD = [f'Gene::{gen}' for gen in lista_id_genes]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "Z5UeD4b0WSj2",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1729202223716,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "Z5UeD4b0WSj2",
    "outputId": "8ba634d5-4877-47f9-c2f3-907d0466e786"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "191"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(lista_genes_EAD)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "H3NOqe4jVtWc",
   "metadata": {
    "id": "H3NOqe4jVtWc"
   },
   "outputs": [],
   "source": [
    "genes_en_grafo = (entidades_genes['Entity_name']).tolist() #Creo lista de los genes presentes en el grafo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "rpf9xut6ghz3",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1729202260474,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "rpf9xut6ghz3",
    "outputId": "2abb4aa5-a8a5-4df8-8524-a9d187173060"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "39220"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(genes_en_grafo) #Cantidad de genes en el grafo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "OVQSm-b-gpYj",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 480,
     "status": "ok",
     "timestamp": 1729202272044,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "OVQSm-b-gpYj",
    "outputId": "2aaf6113-926d-4cda-b312-e7ee6c672ff5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "182\n"
     ]
    }
   ],
   "source": [
    "#Veo cuántos de los genes relacionados a las enfermedades desmielinizantes se encuentran en el grafo:\n",
    "genesEAD = []\n",
    "existe = 0\n",
    "for gen in lista_genes_EAD:\n",
    "  if gen in genes_en_grafo:\n",
    "    existe+=1\n",
    "    genesEAD.append(gen)\n",
    "print(existe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "BodY_RpLhBt_",
   "metadata": {
    "id": "BodY_RpLhBt_"
   },
   "source": [
    "182 de los 191 genes en nuestro listado de genes relacionados con las enfermedades de nuestro interés están presentes en el DRKG, por lo que podemos pensar que los resultados obtenidos al aplicar repurposing con relaciones gen-droga podrían ser aceptables."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2ea1w2A6fgX7",
   "metadata": {
    "id": "2ea1w2A6fgX7"
   },
   "source": [
    "Usamos, al igual que antes, como drogas candidatas aquellas aprobadas por la FDA que se encuentran en Drugbank, excluyendo drogas con peso molecular menor a 250. Esta lista de drogas está en infer_drug.tsv."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "sDGJAGF2d8cv",
   "metadata": {
    "id": "sDGJAGF2d8cv"
   },
   "outputs": [],
   "source": [
    "#Usamos 4 tipos de relaciones entre genes y drogas en forma individual, más el conjunto de las 4 relaciones en forma simultánea:\n",
    "\n",
    "treatment_gen1 = ['GNBR::N::Compound:Gene']\n",
    "treatment_gen2 = ['DRUGBANK::target::Compound:Gene']\n",
    "treatment_gen3 = ['DGIDB::INHIBITOR::Gene:Compound']\n",
    "treatment_gen4 = ['Hetionet::CbG::Compound:Gene']\n",
    "treatment_gen5 = ['GNBR::N::Compound:Gene','DRUGBANK::target::Compound:Gene','DGIDB::INHIBITOR::Gene:Compound','Hetionet::CbG::Compound:Gene']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3-_SuCU0hKnG",
   "metadata": {
    "id": "3-_SuCU0hKnG"
   },
   "outputs": [],
   "source": [
    "# Obtenemos el mapeo de gen o droga a ID de la entidad:\n",
    "entity_map = {}\n",
    "entity_id_map = {}\n",
    "relation_map = {}\n",
    "with open(entity_idmap_file, newline='', encoding='utf-8') as csvfile:\n",
    "    reader = csv.DictReader(csvfile, delimiter='\\t', fieldnames=['name','id'])\n",
    "    for row_val in reader:\n",
    "        entity_map[row_val['name']] = int(row_val['id'])\n",
    "        entity_id_map[int(row_val['id'])] = row_val['name']\n",
    "\n",
    "with open(relation_idmap_file, newline='', encoding='utf-8') as csvfile:\n",
    "    reader = csv.DictReader(csvfile, delimiter='\\t', fieldnames=['name','id'])\n",
    "    for row_val in reader:\n",
    "        relation_map[row_val['name']] = int(row_val['id'])\n",
    "\n",
    "#Obtención de los IDs de drogas, genes y relaciones de tratamientos:\n",
    "drug_ids = []\n",
    "genes_idsEAD = []\n",
    "for drug in drug_list:\n",
    "    drug_ids.append(entity_map[drug])\n",
    "\n",
    "for genes in genesEAD:\n",
    "    genes_idsEAD.append(entity_map[genes])\n",
    "\n",
    "treatment_gen_rid1 = [relation_map[treat]  for treat in treatment_gen1]\n",
    "treatment_gen_rid2 = [relation_map[treat]  for treat in treatment_gen2]\n",
    "treatment_gen_rid3 = [relation_map[treat]  for treat in treatment_gen3]\n",
    "treatment_gen_rid4 = [relation_map[treat]  for treat in treatment_gen4]\n",
    "treatment_gen_rid5 = [relation_map[treat]  for treat in treatment_gen5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "tD_s_qjSkW1g",
   "metadata": {
    "id": "tD_s_qjSkW1g"
   },
   "outputs": [],
   "source": [
    "#Cargamos los embeddings de drogas, genes y tratamientos\n",
    "drug_ids = torch.tensor(drug_ids).long()\n",
    "genes_idsEAD = torch.tensor(genes_idsEAD).long()\n",
    "treatment_gen_rid1 = torch.tensor(treatment_gen_rid1)\n",
    "treatment_gen_rid2 = torch.tensor(treatment_gen_rid2)\n",
    "treatment_gen_rid3 = torch.tensor(treatment_gen_rid3)\n",
    "treatment_gen_rid4 = torch.tensor(treatment_gen_rid4)\n",
    "treatment_gen_rid5 = torch.tensor(treatment_gen_rid5)\n",
    "\n",
    "treatment_gen_embs1 = [torch.tensor(emb_rel[rid]) for rid in treatment_gen_rid1]\n",
    "treatment_gen_embs2 = [torch.tensor(emb_rel[rid]) for rid in treatment_gen_rid2]\n",
    "treatment_gen_embs3 = [torch.tensor(emb_rel[rid]) for rid in treatment_gen_rid3]\n",
    "treatment_gen_embs4 = [torch.tensor(emb_rel[rid]) for rid in treatment_gen_rid4]\n",
    "treatment_gen_embs5 = [torch.tensor(emb_rel[rid]) for rid in treatment_gen_rid5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "S53vioBrkwP5",
   "metadata": {
    "id": "S53vioBrkwP5"
   },
   "outputs": [],
   "source": [
    "gamma=12.0\n",
    "\n",
    "def transE_l2(head, rel, tail):\n",
    "    score = head + rel - tail\n",
    "    return gamma - torch.norm(score, p=2, dim=-1)\n",
    "\n",
    "#Primer relación gen-compuesto:\n",
    "scores_per_gene1 = []\n",
    "dids_per_gene1 = []\n",
    "for rid in range(len(treatment_gen_embs1)):\n",
    "    treatment_emb1=treatment_gen_embs1[rid]\n",
    "    for gene_id in genes_idsEAD:\n",
    "        gene_emb1 = torch.tensor(emb_entity[gene_id])\n",
    "        score1 = fn.logsigmoid(transE_l2(drug_emb, treatment_emb1,\n",
    "                                            gene_emb1))\n",
    "        scores_per_gene1.append(score1)\n",
    "        dids_per_gene1.append(drug_ids)\n",
    "scores1 = torch.cat(scores_per_gene1)\n",
    "dids1 = torch.cat(dids_per_gene1)\n",
    "\n",
    "#Segunda relación gen-compuesto:\n",
    "scores_per_gene2 = []\n",
    "dids_per_gene2 = []\n",
    "for rid in range(len(treatment_gen_embs2)):\n",
    "    treatment_emb2=treatment_gen_embs2[rid]\n",
    "    for gene_id in genes_idsEAD:\n",
    "        gene_emb2 = torch.tensor(emb_entity[gene_id])\n",
    "        score2 = fn.logsigmoid(transE_l2(drug_emb, treatment_emb2,\n",
    "                                            gene_emb2))\n",
    "        scores_per_gene2.append(score2)\n",
    "        dids_per_gene2.append(drug_ids)\n",
    "scores2 = torch.cat(scores_per_gene2)\n",
    "dids2 = torch.cat(dids_per_gene2)\n",
    "\n",
    "#Tercera relación gen-compuesto:\n",
    "scores_per_gene3 = []\n",
    "dids_per_gene3 = []\n",
    "for rid in range(len(treatment_gen_embs3)):\n",
    "    treatment_emb3=treatment_gen_embs3[rid]\n",
    "    for gene_id in genes_idsEAD:\n",
    "        gene_emb3 = torch.tensor(emb_entity[gene_id])\n",
    "        score3 = fn.logsigmoid(transE_l2(gene_emb3, treatment_emb3,\n",
    "                                        drug_emb))\n",
    "        scores_per_gene3.append(score3)\n",
    "        dids_per_gene3.append(drug_ids)\n",
    "scores3 = torch.cat(scores_per_gene3)\n",
    "dids3 = torch.cat(dids_per_gene3)\n",
    "\n",
    "#Cuarta relación gen-compuesto:\n",
    "scores_per_gene4 = []\n",
    "dids_per_gene4 = []\n",
    "for rid in range(len(treatment_gen_embs4)):\n",
    "    treatment_emb4=treatment_gen_embs4[rid]\n",
    "    for gene_id in genes_idsEAD:\n",
    "        gene_emb4 = torch.tensor(emb_entity[gene_id])\n",
    "        score4 = fn.logsigmoid(transE_l2(drug_emb, treatment_emb4,\n",
    "                                            gene_emb4))\n",
    "        scores_per_gene4.append(score4)\n",
    "        dids_per_gene4.append(drug_ids)\n",
    "scores4 = torch.cat(scores_per_gene4)\n",
    "dids4 = torch.cat(dids_per_gene4)\n",
    "\n",
    "#Considerando todas las relaciones gen-compuesto en el DRKG:\n",
    "scores_per_gene5 = []\n",
    "dids_per_gene5 = []\n",
    "for rid in range(len(treatment_gen_embs5)):\n",
    "    treatment_emb5=treatment_gen_embs5[rid]\n",
    "    for gene_id in genes_idsEAD:\n",
    "        gene_emb5 = torch.tensor(emb_entity[gene_id])\n",
    "\n",
    "        if treatment_gen5[rid] == 'DGIDB::INHIBITOR::Gene:Compound':\n",
    "            score5 = fn.logsigmoid(transE_l2(gene_emb5, treatment_emb5,\n",
    "                                        drug_emb))\n",
    "        else:\n",
    "            score5 = fn.logsigmoid(transE_l2(drug_emb, treatment_emb5,\n",
    "                                            gene_emb5))\n",
    "        scores_per_gene5.append(score5)\n",
    "        dids_per_gene5.append(drug_ids)\n",
    "scores5 = torch.cat(scores_per_gene5)\n",
    "dids5 = torch.cat(dids_per_gene5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "BdDvfsOvRloz",
   "metadata": {
    "id": "BdDvfsOvRloz"
   },
   "outputs": [],
   "source": [
    "drogas_ead_file = './drogas_repurposing.tsv' #Cargamos el archivo con las drogas obtenidas en el repurposing a través de relaciones enfermedad-droga:\n",
    "drogas_ead_map = {}\n",
    "with open(drogas_ead_file, newline='', encoding='utf-8') as csvfile:\n",
    "    reader = csv.DictReader(csvfile, delimiter='\\t', fieldnames=['Index', 'Drug Name', 'ID Drugbank'])\n",
    "    for row_val in reader:\n",
    "        drogas_ead_map[row_val['ID Drugbank']] = row_val['Drug Name']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "jn87nXKmlYc0",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 499,
     "status": "ok",
     "timestamp": 1729202470639,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "jn87nXKmlYc0",
    "outputId": "71134a46-5a92-479f-a28c-f81b7fccfe89"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gene::3439\t73\t\n",
      "[0]Tretinoin\n",
      "[1]Thalidomide\n",
      "[2]Methotrexate\n",
      "[3]Dexamethasone\n",
      "[4]Zidovudine\n",
      "[5]Ribavirin\n",
      "[6]Paclitaxel\n",
      "[7]Cisplatin\n",
      "[8]Gemcitabine\n",
      "[9]Doxorubicin\n",
      "[10]Hydrocortisone\n",
      "[11]Sirolimus\n",
      "[12]Prednisone\n",
      "[13]Fludarabine\n",
      "[14]Methylprednisolone\n",
      "[15]Cyclosporine\n",
      "[16]Chloroquine\n",
      "[17]Etoposide\n",
      "[18]Topotecan\n",
      "[19]Prednisolone\n",
      "[20]Vincristine\n",
      "[21]Isotretinoin\n",
      "[22]Azathioprine\n",
      "[23]Curcumin\n",
      "[24]Dinoprostone\n",
      "[25]Hydroxychloroquine\n",
      "[27]Carboplatin\n",
      "[28]Budesonide\n",
      "[29]Imatinib\n",
      "[30]Colchicine\n",
      "[31]Estradiol\n",
      "[32]Mitoxantrone\n",
      "[33]Dactinomycin\n",
      "[35]Sunitinib\n",
      "[36]Quercetin\n",
      "[37]Testosterone\n",
      "[40]Cyclophosphamide\n",
      "[41]Pazopanib\n",
      "[43]Dasatinib\n",
      "[44]Genistein\n",
      "[45]Sulfamethoxazole\n",
      "[46]Indomethacin\n",
      "[47]Melatonin\n",
      "[48]DB12116\n",
      "[49]Pentoxifylline\n",
      "[50]Cladribine\n",
      "[51]Mycophenolate mofetil\n",
      "[54]Vinblastine\n",
      "[55]Simvastatin\n",
      "[56]Mitomycin\n",
      "[57]Sulfasalazine\n",
      "[58]Idarubicin\n",
      "[59]Roflumilast\n",
      "[64]Bleomycin\n",
      "[65]Vitamin D\n",
      "[66]Morphine\n",
      "[67]Progesterone\n",
      "[68]Tacrolimus\n",
      "[69]Irinotecan\n",
      "[72]Epirubicin\n",
      "[74]Daunorubicin\n",
      "[76]Fenofibrate\n",
      "[81]Nilotinib\n",
      "[83]Cyclic Adenosine Monophosphate\n",
      "[85]Adenosine\n",
      "[86]Mycophenolic acid\n",
      "[88]Fluticasone Propionate\n",
      "[89]Itraconazole\n",
      "[90]Diazepam\n",
      "[91]Betamethasone\n",
      "[92]Cholesterol\n",
      "[96]Naloxone\n",
      "[97]Metoclopramide\n",
      "\n",
      "DB01229\tPaclitaxel\t180\t18.69970117195758\n",
      "DB01234\tDexamethasone\t179\t27.82749118980998\n",
      "DB00877\tSirolimus\t179\t18.10447245728168\n",
      "DB00755\tTretinoin\t178\t27.42267752030995\n",
      "DB04540\tCholesterol\t178\t23.39788348975694\n",
      "DB01645\tGenistein\t177\t13.598333834120712\n",
      "DB00997\tDoxorubicin\t175\t12.507110606199644\n",
      "DB00515\tCisplatin\t173\t27.63749782961143\n",
      "DB11672\tCurcumin\t173\t23.61048106026915\n",
      "DB01065\tMelatonin\t172\t12.590252833414056\n",
      "DB00783\tEstradiol\t172\t7.897086584748297\n",
      "DB09341\tGlucose\t171\t10.154873653106346\n",
      "DB04216\tQuercetin\t171\t12.750423242952355\n",
      "DB00563\tMethotrexate\t169\t19.484590802995058\n",
      "DB00441\tGemcitabine\t164\t9.39879052675372\n",
      "DB01041\tThalidomide\t162\t9.743987230189822\n",
      "DB00641\tSimvastatin\t161\t10.053510257705712\n",
      "DB12116\tDB12116\t157\t27.619252337957807\n",
      "DB00741\tHydrocortisone\t150\t12.203702007131591\n",
      "DB00396\tProgesterone\t150\t5.00428627134673\n",
      "DB00091\tCyclosporine\t148\t16.700196766510622\n",
      "DB02527\tCyclic Adenosine Monophosphate\t146\t9.366792593830127\n",
      "DB11094\tVitamin D\t139\t8.999122882839739\n",
      "DB00328\tIndomethacin\t139\t6.8324282944416765\n",
      "DB00970\tDactinomycin\t139\t3.2732969054576366\n",
      "DB00917\tDinoprostone\t137\t13.13258276630953\n",
      "DB00640\tAdenosine\t134\t6.849678927910629\n",
      "DB00958\tCarboplatin\t126\t3.578861654405161\n",
      "DB00773\tEtoposide\t124\t2.7992465938660067\n",
      "DB04630\tAldosterone\t120\t2.33820421339294\n",
      "DB00608\tChloroquine\t116\t2.718923061233452\n",
      "DB00795\tSulfasalazine\t116\t3.3508898196685806\n",
      "DB00982\tIsotretinoin\t115\t2.9788095042917853\n",
      "DB04115\tBerberine\t114\t5.817479453259252\n",
      "DB00959\tMethylprednisolone\t113\t4.959108142474747\n",
      "DB00305\tMitomycin\t112\t2.581412158269812\n",
      "DB00635\tPrednisone\t111\t4.47122926243533\n",
      "DB00624\tTestosterone\t111\t3.0513839015498254\n",
      "DB00806\tPentoxifylline\t108\t3.1879664228683353\n",
      "DB06774\tDB06774\t107\t3.671147797675071\n",
      "DB06589\tPazopanib\t107\t2.6288382619881676\n",
      "DB00860\tPrednisolone\t99\t3.6405354790088684\n",
      "DB00541\tVincristine\t97\t2.20793456663687\n",
      "DB01222\tBudesonide\t95\t4.272382456305731\n",
      "DB01394\tColchicine\t91\t1.6238490281474718\n",
      "DB01039\tFenofibrate\t91\t2.497840834926212\n",
      "DB01204\tMitoxantrone\t90\t2.281708553600416\n",
      "DB01268\tSunitinib\t89\t1.9679563806761629\n",
      "DB00811\tRibavirin\t77\t2.9460779425981056\n",
      "DB01073\tFludarabine\t76\t2.1449115193670045\n",
      "DB00993\tAzathioprine\t73\t1.4044479809181136\n",
      "DB00495\tZidovudine\t72\t2.570287760638307\n",
      "DB00158\tFolic Acid\t68\t1.7711551552661213\n",
      "DB01092\tOuabain\t67\t1.8279835902485728\n",
      "DB01254\tDasatinib\t64\t2.076506746628664\n",
      "DB00694\tDaunorubicin\t62\t1.1510778264163544\n",
      "DB01030\tTopotecan\t61\t1.0636034128096623\n",
      "DB00975\tDipyridamole\t61\t1.2834947316896663\n",
      "DB01103\tQuinacrine\t59\t1.362840104744665\n",
      "DB00295\tMorphine\t56\t1.4176724136793761\n",
      "DB01118\tAmiodarone\t56\t0.7914777982851589\n",
      "DB00688\tMycophenolate mofetil\t55\t0.8276318631061351\n",
      "DB00136\tCalcitriol\t54\t1.0768840102072044\n",
      "DB00999\tHydrochlorothiazide\t52\t1.001099673703574\n",
      "DB00908\tQuinidine\t49\t0.9578206549447992\n",
      "DB00570\tVinblastine\t48\t1.021255519524796\n",
      "DB00143\tGlutathione\t47\t1.9370196398083945\n",
      "DB01183\tNaloxone\t46\t1.3382438172974829\n",
      "DB01611\tHydroxychloroquine\t45\t0.7939284284894559\n",
      "DB01211\tClarithromycin\t44\t0.8146598424275852\n",
      "DB01233\tMetoclopramide\t41\t0.6821492603674487\n",
      "DB01017\tMinocycline\t39\t0.6841715825283661\n",
      "DB00619\tImatinib\t39\t1.2607584372707612\n",
      "DB04868\tNilotinib\t38\t0.5571715686023058\n",
      "DB00445\tEpirubicin\t35\t0.5422315631922217\n",
      "DB00502\tHaloperidol\t30\t0.9584685065050501\n",
      "DB00822\tDisulfiram\t27\t0.5182910634258601\n",
      "DB01167\tItraconazole\t27\t0.3519958785443354\n",
      "DB01656\tRoflumilast\t27\t0.3884965185843815\n",
      "DB01248\tDocetaxel\t27\t0.3925278532253814\n",
      "DB00443\tBetamethasone\t26\t0.3934016308649714\n",
      "DB01045\tRifampicin\t26\t0.5175156970316074\n",
      "DB00882\tClomifene\t26\t0.42362253671826133\n",
      "DB14128\tNadide\t25\t4.066471701501989\n",
      "DB00762\tIrinotecan\t24\t0.39015510187682373\n",
      "DB00829\tDiazepam\t24\t1.3962828393728035\n",
      "DB01165\tOfloxacin\t23\t0.30679693988537177\n",
      "DB00242\tCladribine\t22\t0.3391797438577941\n",
      "DB00169\tCholecalciferol\t20\t0.839714392881894\n",
      "DB14104\tLinoleic acid\t20\t0.3333116659938735\n",
      "DB00864\tTacrolimus\t19\t0.25714646301138644\n",
      "DB00620\tTriamcinolone\t18\t0.2691636309791836\n",
      "DB01015\tSulfamethoxazole\t17\t0.27407138416188415\n",
      "DB01241\tGemfibrozil\t16\t0.34265540395619193\n",
      "DB01200\tBromocriptine\t15\t0.5460966133717349\n",
      "DB00472\tFluoxetine\t15\t0.35787113664457537\n",
      "DB01177\tIdarubicin\t15\t0.2156625786073036\n",
      "DB01024\tMycophenolic acid\t13\t0.1887416332924225\n",
      "DB00470\tDronabinol\t12\t0.9995113004391538\n",
      "DB00904\tOndansetron\t11\t0.25145166161957205\n",
      "DB00591\tFluocinolone Acetonide\t10\t0.14228750057859418\n",
      "DB08910\tPomalidomide\t10\t0.12037425034961455\n",
      "DB00290\tBleomycin\t9\t0.2128614892436924\n",
      "DB08880\tTeriflunomide\t8\t0.11019173190663724\n",
      "DB01042\tMelphalan\t7\t0.10894435824181554\n",
      "DB11526\tDB11526\t7\t0.37006789933711387\n",
      "DB01281\tAbatacept\t6\t0.3027242515721778\n",
      "DB01026\tKetoconazole\t6\t0.15269588910156046\n",
      "DB00977\tEthinyl Estradiol\t6\t0.09377415306965207\n",
      "DB01104\tSertraline\t6\t0.0946795521754344\n",
      "DB00642\tPemetrexed\t4\t0.04900815243331996\n",
      "DB01137\tLevofloxacin\t4\t0.06827117327117327\n",
      "DB01166\tCilostazol\t3\t0.03971161114018257\n",
      "DB00531\tCyclophosphamide\t3\t0.05412922550278135\n",
      "DB00552\tPentostatin\t3\t0.04457243767400061\n",
      "DB00588\tFluticasone Propionate\t3\t0.03592731308087113\n",
      "DB01097\tLeflunomide\t2\t0.021980676328502417\n",
      "DB00068\tInterferon beta-1b\t2\t0.15601503759398494\n",
      "DB00022\tPeginterferon alfa-2b\t2\t1.0384615384615385\n",
      "DB01068\tClonazepam\t2\t0.34444444444444444\n",
      "DB05459\tABT-874\t2\t0.024501173708920188\n",
      "DB04574\tEstropipate\t2\t0.0529616724738676\n",
      "DB00105\tInterferon Alfa-2b, Recombinant\t1\t0.04\n",
      "DB00008\tPeginterferon alfa-2a\t1\t0.05263157894736842\n",
      "DB00033\tInterferon gamma-1b\t1\t0.16666666666666666\n",
      "DB00074\tBasiliximab\t1\t0.01098901098901099\n",
      "DB13038\tGlyceryl trioleate\t1\t0.01694915254237288\n",
      "DB00108\tNatalizumab\t1\t0.019230769230769232\n",
      "DB09033\tDB09033\t1\t0.03225806451612903\n",
      "DB01380\tCortisone acetate\t1\t0.011235955056179775\n",
      "DB01285\tCorticotropin\t1\t0.014084507042253521\n",
      "DB01108\tTrilostane\t1\t0.02\n",
      "DB08904\tCertolizumab pegol\t1\t0.047619047619047616\n"
     ]
    }
   ],
   "source": [
    "maxhit1=0 #número máximo de \"hits\" (coincidencias) entre genes y drogas\n",
    "drugs_in_top_k1={} #cuántas veces una droga aparece en el top de las más relacionadas con un gen\n",
    "drugsfr_in_top_k1={}\n",
    "for i in range(len(scores_per_gene1)):\n",
    "    score_g1=scores_per_gene1[i] #lista con los puntajes de las drogas asociadas al gen\n",
    "    did_g1=dids_per_gene1[i] #lista con los IDs de las drogas asociadas a ese gen\n",
    "    idx_g1 = torch.flip(torch.argsort(score_g1), dims=[0]) #Ordena los índices de los elementos del tensor de mayor a menor\n",
    "    score_g1 = score_g1[idx_g1].numpy() #se ordenan los scores de mayor a menor\n",
    "    did_g1 = did_g1[idx_g1].numpy() #se ordenan los IDs de las drogas en función de sus scores de mayor a menor\n",
    "    #print(did)\n",
    "    _, unique_indices1 = np.unique(did_g1, return_index=True) #Devuelve los índices donde aparecen por primera vez los valores únicos en did_g\n",
    "    topk=100\n",
    "    topk_indices1 = np.sort(unique_indices1)[:topk]  # Da los primeros 100 índices (o menos, si no hay 100) de los valores únicos en did_g, ordenados en forma ascendente.\n",
    "    proposed_did_g1 = did_g1[topk_indices1] #primeras 100 drogas\n",
    "    proposed_score_g1 = score_g1[topk_indices1] #primeros 100 scores\n",
    "    found_in_top_k1=0\n",
    "    found_drugs1=\"\\n\"\n",
    "    for j in range(topk): #loop para buscar en la lista de las 100 mejores drogas si alguna de ellas está en el diccionario drogas_ead_map (drogas que obtuvimos del repurposing con relación enfermedad-droga)\n",
    "        drug_g1 = entity_id_map[int(proposed_did_g1[j])][10:17]\n",
    "        if drogas_ead_map.get(drug_g1, None) is not None:\n",
    "            found_in_top_k1+=1 #si la droga está en el dicionario, se incrementa este contador\n",
    "            score1 = proposed_score_g1[j]\n",
    "            if drug_g1 in drugs_in_top_k1:\n",
    "                drugs_in_top_k1[drug_g1]+=1\n",
    "                drugsfr_in_top_k1[drug_g1]+=1/(j+1) #1/(j+1) valor ponderado para la droga según sus posición en ranking para un gen dado (se van sumando)\n",
    "            else:\n",
    "                drugs_in_top_k1[drug_g1]=1\n",
    "                drugsfr_in_top_k1[drug_g1]=1/(j+1)\n",
    "            found_drugs1+=\"[{}]{}\\n\".format(j, drogas_ead_map[drug_g1]) #almacenamiento del nombre de la droga\n",
    "            #print(\"[{}]{}\".format(j, clinical_drug_map[drug]))\n",
    "    #print(\"{}\\t{}\".format(cov_related_genes[i], found_in_top_k))\n",
    "    if maxhit1< found_in_top_k1:\n",
    "        maxhit1=found_in_top_k1\n",
    "        maxgene1=genesEAD[i]\n",
    "        max_dugs1=found_drugs1\n",
    "print(\"{}\\t{}\\t{}\".format(maxgene1, maxhit1,max_dugs1)) #Se imprime el gen de los relacionados con las EADs con más hits de los medicamentos propuestos para dichas enfermedades,\n",
    "                                                     #el número de hits para ese gen, y los nombres de los medicamentos asociados al mismo.\n",
    "\n",
    "res1=[[drug, drogas_ead_map[drug] ,drugs_in_top_k1[drug],drugsfr_in_top_k1[drug]] for drug in drugs_in_top_k1.keys()]\n",
    "res1=reversed(sorted(res1, key=lambda x : x[2]))\n",
    "for drug in res1:\n",
    "    print(\"{}\\t{}\\t{}\\t{}\".format(drug[0], drug[1] ,drug[2],drug[3])) #Imprime un listado con IDs y nombres de las drogas que actúan sobre más genes, con el nro de veces\n",
    "                                                                      #que aparece la droga en el top 100, y un valor ponderado por su ranking (\n",
    "                                                                      #su mayor valor sería en caso que la droga aparezca primera en todos los rankings de los 182 genes, o sea,\n",
    "                                                                      #sería igual a 182/2 = 91, mientras que el peor valor es si aparece última y en una sola lista, o sea, 1/100 = 0.01\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "SkqACDYuifq-",
   "metadata": {
    "id": "SkqACDYuifq-"
   },
   "source": [
    "El gen con más hits, o relacionado con mayor cantidad dentro de los medicamentos analizados es el que posee ID Entrez 3439, o sea, IFNA1 ó interferon alpha 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "WeJZNZ6VOYOq",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 509,
     "status": "ok",
     "timestamp": 1729202477959,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "WeJZNZ6VOYOq",
    "outputId": "a140f9a2-848f-4aaf-ee4d-2689bb985fd8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gene::3439\t56\t\n",
      "[2]Sunitinib\n",
      "[3]Zidovudine\n",
      "[4]Imatinib\n",
      "[6]Tretinoin\n",
      "[7]Dexamethasone\n",
      "[8]Thalidomide\n",
      "[10]Cladribine\n",
      "[11]Dasatinib\n",
      "[14]Methotrexate\n",
      "[15]Nilotinib\n",
      "[16]Topotecan\n",
      "[18]Doxorubicin\n",
      "[20]Azathioprine\n",
      "[21]Hydroxychloroquine\n",
      "[22]Cyclophosphamide\n",
      "[23]Pazopanib\n",
      "[24]Ribavirin\n",
      "[25]Cisplatin\n",
      "[26]Fludarabine\n",
      "[29]Cyclosporine\n",
      "[31]Prednisone\n",
      "[32]Estradiol\n",
      "[33]Sulfamethoxazole\n",
      "[35]Paclitaxel\n",
      "[37]Etoposide\n",
      "[38]Vinblastine\n",
      "[40]Fluticasone Propionate\n",
      "[41]Colchicine\n",
      "[44]ABT-874\n",
      "[47]Sirolimus\n",
      "[48]Tacrolimus\n",
      "[49]Irinotecan\n",
      "[50]Basiliximab\n",
      "[51]Tocilizumab\n",
      "[52]Methylprednisolone\n",
      "[54]Vincristine\n",
      "[57]Mycophenolate mofetil\n",
      "[58]Budesonide\n",
      "[61]Idarubicin\n",
      "[62]Docetaxel\n",
      "[63]Hydrocortisone\n",
      "[64]Roflumilast\n",
      "[66]Gemcitabine\n",
      "[69]Mitoxantrone\n",
      "[74]Chloroquine\n",
      "[77]Certolizumab pegol\n",
      "[78]Carboplatin\n",
      "[80]Indomethacin\n",
      "[83]Leflunomide\n",
      "[84]Prednisolone\n",
      "[88]Quercetin\n",
      "[89]Bleomycin\n",
      "[90]Dactinomycin\n",
      "[94]Mycophenolic acid\n",
      "[95]Estropipate\n",
      "[96]Testosterone\n",
      "\n",
      "DB00783\tEstradiol\t167\t10.883496556552092\n",
      "DB01234\tDexamethasone\t160\t13.911560039722648\n",
      "DB01268\tSunitinib\t153\t19.63300285242422\n",
      "DB00755\tTretinoin\t149\t8.989551392193004\n",
      "DB06589\tPazopanib\t143\t5.501893657219361\n",
      "DB00997\tDoxorubicin\t140\t6.215505688776108\n",
      "DB00515\tCisplatin\t138\t10.754680700408052\n",
      "DB01254\tDasatinib\t129\t4.2981982411370785\n",
      "DB04216\tQuercetin\t128\t8.765073919098203\n",
      "DB00877\tSirolimus\t117\t3.5415171859971113\n",
      "DB01229\tPaclitaxel\t116\t3.3938686761221946\n",
      "DB00091\tCyclosporine\t113\t8.838809759729571\n",
      "DB04868\tNilotinib\t112\t5.772726423097177\n",
      "DB00328\tIndomethacin\t108\t7.391376131124137\n",
      "DB00563\tMethotrexate\t106\t4.583183499413917\n",
      "DB01065\tMelatonin\t99\t4.97412452643374\n",
      "DB11672\tCurcumin\t94\t3.7197217511165044\n",
      "DB04574\tEstropipate\t87\t2.2285070134105642\n",
      "DB00570\tVinblastine\t86\t2.6962435451851157\n",
      "DB06774\tDB06774\t85\t2.90829625775966\n",
      "DB01645\tGenistein\t79\t3.9948123644364095\n",
      "DB08904\tCertolizumab pegol\t78\t2.2033274555812716\n",
      "DB00619\tImatinib\t76\t3.86793463740168\n",
      "DB01041\tThalidomide\t74\t2.3568467096222863\n",
      "DB00795\tSulfasalazine\t73\t2.0306056903169445\n",
      "DB08910\tPomalidomide\t65\t1.3532965514929314\n",
      "DB00286\tConjugated Estrogens\t65\t1.3128386155017557\n",
      "DB00242\tCladribine\t65\t2.022500760704199\n",
      "DB01248\tDocetaxel\t65\t1.6816174381682976\n",
      "DB01394\tColchicine\t61\t1.2331403896325908\n",
      "DB00688\tMycophenolate mofetil\t60\t1.1115941966029403\n",
      "DB01017\tMinocycline\t58\t1.8148856724522526\n",
      "DB00993\tAzathioprine\t54\t1.1602796364647896\n",
      "DB00977\tEthinyl Estradiol\t54\t0.9346374625429316\n",
      "DB01222\tBudesonide\t51\t2.2588225633941286\n",
      "DB00959\tMethylprednisolone\t50\t1.353911154169589\n",
      "DB00635\tPrednisone\t49\t1.3877627222177789\n",
      "DB00741\tHydrocortisone\t47\t1.2594796087593405\n",
      "DB00495\tZidovudine\t47\t1.5589178145293914\n",
      "DB01092\tOuabain\t45\t2.158140141715916\n",
      "DB00624\tTestosterone\t45\t1.0153717313278694\n",
      "DB00773\tEtoposide\t44\t0.9798494785714246\n",
      "DB11094\tVitamin D\t43\t0.940974153600879\n",
      "DB00975\tDipyridamole\t40\t1.154134403182763\n",
      "DB00970\tDactinomycin\t39\t0.5918933538806498\n",
      "DB01611\tHydroxychloroquine\t38\t0.7737648667135597\n",
      "DB00864\tTacrolimus\t38\t0.7350788151018156\n",
      "DB01030\tTopotecan\t37\t0.8240038292733239\n",
      "DB00762\tIrinotecan\t37\t1.0196249807322875\n",
      "DB01166\tCilostazol\t36\t0.6752996275380626\n",
      "DB00958\tCarboplatin\t35\t0.7505180956755269\n",
      "DB00143\tGlutathione\t35\t1.0780433398523608\n",
      "DB00860\tPrednisolone\t34\t0.6540477450525828\n",
      "DB01015\tSulfamethoxazole\t32\t0.619569028190274\n",
      "DB00441\tGemcitabine\t32\t0.5801841481374336\n",
      "DB00917\tDinoprostone\t32\t1.2350642503773257\n",
      "DB00396\tProgesterone\t31\t0.7607659278232249\n",
      "DB00694\tDaunorubicin\t31\t0.555063472917541\n",
      "DB00908\tQuinidine\t30\t0.7263596168467646\n",
      "DB00806\tPentoxifylline\t29\t0.6698844674775631\n",
      "DB06273\tTocilizumab\t29\t0.5872956350203075\n",
      "DB01183\tNaloxone\t28\t1.0281554767013008\n",
      "DB00295\tMorphine\t27\t0.7202040522559989\n",
      "DB00502\tHaloperidol\t27\t1.0895961181285465\n",
      "DB00588\tFluticasone Propionate\t26\t0.5589627776116327\n",
      "DB01118\tAmiodarone\t26\t0.41504049196411136\n",
      "DB01073\tFludarabine\t24\t0.562789379522392\n",
      "DB01656\tRoflumilast\t23\t0.37816580938193256\n",
      "DB01204\tMitoxantrone\t23\t0.4474134026666584\n",
      "DB00641\tSimvastatin\t22\t0.705782438252277\n",
      "DB08880\tTeriflunomide\t21\t0.3697803436778211\n",
      "DB01039\tFenofibrate\t20\t0.892700826553622\n",
      "DB00541\tVincristine\t20\t0.3827037347821985\n",
      "DB00591\tFluocinolone Acetonide\t19\t0.3064168538019169\n",
      "DB04540\tCholesterol\t19\t2.967586356574285\n",
      "DB00620\tTriamcinolone\t18\t0.36886195114431064\n",
      "DB01200\tBromocriptine\t15\t0.7451110184857334\n",
      "DB00443\tBetamethasone\t15\t0.28725886804264406\n",
      "DB00074\tBasiliximab\t14\t0.858071392652718\n",
      "DB01103\tQuinacrine\t14\t0.21797964508802878\n",
      "DB00472\tFluoxetine\t13\t0.3957939188219422\n",
      "DB01024\tMycophenolic acid\t12\t0.2077550916913867\n",
      "DB01233\tMetoclopramide\t12\t0.22170765639606912\n",
      "DB01177\tIdarubicin\t12\t0.1997447296995744\n",
      "DB06769\tDB06769\t12\t0.2686392646150942\n",
      "DB04630\tAldosterone\t12\t0.2007299060114577\n",
      "DB00608\tChloroquine\t11\t0.1554833104038556\n",
      "DB00811\tRibavirin\t11\t0.5202092970486027\n",
      "DB00640\tAdenosine\t10\t0.15519465835839563\n",
      "DB00108\tNatalizumab\t10\t0.45914638193855734\n",
      "DB00531\tCyclophosphamide\t10\t0.1782656656643875\n",
      "DB02527\tCyclic Adenosine Monophosphate\t10\t0.15338017530140954\n",
      "DB00999\tHydrochlorothiazide\t10\t0.21803409854781397\n",
      "DB04956\tAfelimomab\t9\t0.1290812590188817\n",
      "DB00158\tFolic Acid\t8\t0.13841363185235686\n",
      "DB09341\tGlucose\t8\t0.19068320649692372\n",
      "DB00982\tIsotretinoin\t7\t0.11441480465970151\n",
      "DB01281\tAbatacept\t7\t0.6571394937086774\n",
      "DB00642\tPemetrexed\t7\t0.1321467747938336\n",
      "DB00290\tBleomycin\t7\t0.1476101429499747\n",
      "DB01097\tLeflunomide\t7\t0.09651216867341642\n",
      "DB01137\tLevofloxacin\t6\t0.13608298510661096\n",
      "DB01241\tGemfibrozil\t6\t0.2003348369407275\n",
      "DB12116\tDB12116\t6\t0.2089618519195984\n",
      "DB09033\tDB09033\t6\t0.27049480924480923\n",
      "DB00445\tEpirubicin\t6\t0.08286318035986324\n",
      "DB00305\tMitomycin\t5\t0.061756623990666545\n",
      "DB01045\tRifampicin\t5\t0.2023115773115773\n",
      "DB01410\tCiclesonide\t5\t0.07586579912925344\n",
      "DB01042\tMelphalan\t5\t0.06979029929216092\n",
      "DB11526\tDB11526\t4\t0.08845896031250723\n",
      "DB01211\tClarithromycin\t4\t0.0452192423361097\n",
      "DB01380\tCortisone acetate\t4\t0.06527777777777778\n",
      "DB05459\tABT-874\t4\t0.4335978835978836\n",
      "DB01104\tSertraline\t3\t0.06746031746031746\n",
      "DB00068\tInterferon beta-1b\t3\t0.6770833333333333\n",
      "DB00022\tPeginterferon alfa-2b\t3\t0.6805555555555555\n",
      "DB00822\tDisulfiram\t3\t0.041262148214547975\n",
      "DB01384\tParamethasone\t3\t0.05210409745293466\n",
      "DB01285\tCorticotropin\t3\t0.07342522446689113\n",
      "DB05259\tGlatiramer Acetate\t3\t0.032377428307123035\n",
      "DB00882\tClomifene\t2\t0.037836147592245156\n",
      "DB14783\tDiroximel fumarate\t2\t0.05505952380952381\n",
      "DB00105\tInterferon Alfa-2b, Recombinant\t2\t0.16993464052287582\n",
      "DB00008\tPeginterferon alfa-2a\t2\t0.26666666666666666\n",
      "DB00033\tInterferon gamma-1b\t2\t1.1666666666666667\n",
      "DB00470\tDronabinol\t2\t0.07573632538569425\n",
      "DB14128\tNadide\t2\t0.5833333333333334\n",
      "DB00552\tPentostatin\t2\t0.021910919540229883\n",
      "DB00829\tDiazepam\t2\t1.0151515151515151\n",
      "DB00904\tOndansetron\t2\t0.02254305977710233\n",
      "DB01165\tOfloxacin\t2\t0.04316239316239316\n",
      "DB01026\tKetoconazole\t2\t0.08319327731092437\n",
      "DB00169\tCholecalciferol\t2\t0.2131578947368421\n",
      "DB01068\tClonazepam\t1\t0.5\n",
      "DB00764\tMometasone\t1\t0.0136986301369863\n",
      "DB01223\tAminophylline\t1\t0.015873015873015872\n",
      "DB00601\tLinezolid\t1\t0.013157894736842105\n",
      "DB01108\tTrilostane\t1\t0.02631578947368421\n",
      "DB00136\tCalcitriol\t1\t0.125\n",
      "DB11988\tDB11988\t1\t0.010101010101010102\n",
      "DB09122\tPeginterferon beta-1a\t1\t0.010638297872340425\n"
     ]
    }
   ],
   "source": [
    "maxhit2=0\n",
    "drugs_in_top_k2={}\n",
    "drugsfr_in_top_k2={}\n",
    "for i in range(len(scores_per_gene2)):\n",
    "    score_g2=scores_per_gene2[i]\n",
    "    did_g2 = dids_per_gene2[i]\n",
    "    idx_g2 = torch.flip(torch.argsort(score_g2), dims=[0]) #Ordena los índices de los elementos del tensor ordenados de mayor a menor\n",
    "    score_g2 = score_g2[idx_g2].numpy()\n",
    "    did_g2 = did_g2[idx_g2].numpy()\n",
    "    #print(did)\n",
    "    _, unique_indices2 = np.unique(did_g2, return_index=True) #Devuelve los índices donde aparecen por primera vez los valores únicos en did_g\n",
    "    topk=100\n",
    "    topk_indices2 = np.sort(unique_indices2)[:topk]  # Da los primeros 100 índices (o menos, si no hay 100) de los valores únicos en did_g, ordenados en forma ascendente.\n",
    "    proposed_did_g2 = did_g2[topk_indices2] #primeras 100 drogas\n",
    "    proposed_score_g2 = score_g2[topk_indices2] #primeros 100 scores\n",
    "    found_in_top_k2=0\n",
    "    found_drugs2=\"\\n\"\n",
    "    for j in range(topk):\n",
    "        drug_g2 = entity_id_map[int(proposed_did_g2[j])][10:17]\n",
    "        if drogas_ead_map.get(drug_g2, None) is not None:\n",
    "            found_in_top_k2+=1\n",
    "            score2 = proposed_score_g2[j]\n",
    "            if drug_g2 in drugs_in_top_k2:\n",
    "                drugs_in_top_k2[drug_g2]+=1\n",
    "                drugsfr_in_top_k2[drug_g2]+=1/(j+1)\n",
    "            else:\n",
    "                drugs_in_top_k2[drug_g2]=1\n",
    "                drugsfr_in_top_k2[drug_g2]=1/(j+1)\n",
    "            found_drugs2+=\"[{}]{}\\n\".format(j, drogas_ead_map[drug_g2])\n",
    "            #print(\"[{}]{}\".format(j, clinical_drug_map[drug]))\n",
    "    #print(\"{}\\t{}\".format(cov_related_genes[i], found_in_top_k))\n",
    "    if maxhit2< found_in_top_k2:\n",
    "        maxhit2=found_in_top_k2\n",
    "        maxgene2=genesEAD[i]\n",
    "        max_dugs2=found_drugs2\n",
    "print(\"{}\\t{}\\t{}\".format(maxgene2, maxhit2,max_dugs2)) #Se imprime el gen de los relacionados con las EADs con más hits de los medicamentos propuestos para dichas enfermedades,\n",
    "                                                     #el número de hits, y los nombres de los medicamentos asociados al gen.\n",
    "\n",
    "res2=[[drug, drogas_ead_map[drug] ,drugs_in_top_k2[drug],drugsfr_in_top_k2[drug]] for drug in drugs_in_top_k2.keys()]\n",
    "res2=reversed(sorted(res2, key=lambda x : x[2]))\n",
    "for drug in res2:\n",
    "    print(\"{}\\t{}\\t{}\\t{}\".format(drug[0], drug[1] ,drug[2],drug[3]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "qInOunXhxu2n",
   "metadata": {
    "id": "qInOunXhxu2n"
   },
   "source": [
    "El gen con más hits, o relacionado con mayor cantidad dentro de los medicamentos analizados es el que posee ID Entrez 3439, o sea, IFNA1 ó interferon alpha 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "61xQ5S1WOYCE",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 493,
     "status": "ok",
     "timestamp": 1729202488797,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "61xQ5S1WOYCE",
    "outputId": "87dc7871-1ce3-4cae-8186-a9733967f4b4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gene::3439\t58\t\n",
      "[0]Zidovudine\n",
      "[1]Sunitinib\n",
      "[2]Tretinoin\n",
      "[3]Hydroxychloroquine\n",
      "[4]Imatinib\n",
      "[5]Dexamethasone\n",
      "[6]Cyclophosphamide\n",
      "[7]Thalidomide\n",
      "[8]Paclitaxel\n",
      "[10]Dasatinib\n",
      "[12]Topotecan\n",
      "[13]Nilotinib\n",
      "[14]Fluticasone Propionate\n",
      "[16]Methotrexate\n",
      "[17]Tacrolimus\n",
      "[22]Etoposide\n",
      "[23]Mitoxantrone\n",
      "[25]Pazopanib\n",
      "[27]Ribavirin\n",
      "[28]Azathioprine\n",
      "[29]Irinotecan\n",
      "[30]Prednisone\n",
      "[31]Fludarabine\n",
      "[32]Docetaxel\n",
      "[34]Methylprednisolone\n",
      "[36]Dactinomycin\n",
      "[37]Hydrocortisone\n",
      "[38]Sirolimus\n",
      "[39]Doxorubicin\n",
      "[40]Idarubicin\n",
      "[41]Cyclosporine\n",
      "[46]Chloroquine\n",
      "[49]Morphine\n",
      "[51]Cladribine\n",
      "[53]Sulfamethoxazole\n",
      "[54]Cisplatin\n",
      "[55]Vincristine\n",
      "[56]Leflunomide\n",
      "[57]Budesonide\n",
      "[58]Gemcitabine\n",
      "[60]Mycophenolate mofetil\n",
      "[61]Quercetin\n",
      "[63]Melphalan\n",
      "[65]Prednisolone\n",
      "[66]Indomethacin\n",
      "[68]Pentoxifylline\n",
      "[69]Carboplatin\n",
      "[70]Colchicine\n",
      "[78]Naloxone\n",
      "[80]Testosterone\n",
      "[82]Bromocriptine\n",
      "[86]Mycophenolic acid\n",
      "[87]Vinblastine\n",
      "[89]Metoclopramide\n",
      "[90]Estradiol\n",
      "[93]Sulfasalazine\n",
      "[96]Betamethasone\n",
      "[97]Daunorubicin\n",
      "\n",
      "DB01234\tDexamethasone\t177\t30.573165026816003\n",
      "DB00755\tTretinoin\t172\t20.544962576659746\n",
      "DB01229\tPaclitaxel\t169\t20.420194571785274\n",
      "DB04216\tQuercetin\t159\t19.497961219020805\n",
      "DB01268\tSunitinib\t156\t20.774943349759383\n",
      "DB06589\tPazopanib\t155\t7.423884652737754\n",
      "DB00877\tSirolimus\t153\t5.794082251798429\n",
      "DB01254\tDasatinib\t149\t8.42172331293365\n",
      "DB04868\tNilotinib\t141\t9.90309460057732\n",
      "DB00970\tDactinomycin\t135\t4.267260989248139\n",
      "DB00328\tIndomethacin\t134\t9.196350121528518\n",
      "DB01248\tDocetaxel\t132\t6.444334164682465\n",
      "DB00563\tMethotrexate\t129\t6.7767692794094\n",
      "DB00864\tTacrolimus\t127\t3.9285721830271187\n",
      "DB00997\tDoxorubicin\t125\t3.5858383209783016\n",
      "DB00783\tEstradiol\t121\t4.00753439138275\n",
      "DB00515\tCisplatin\t121\t4.866626603269444\n",
      "DB01204\tMitoxantrone\t120\t5.005306485318714\n",
      "DB01041\tThalidomide\t116\t5.007276882443757\n",
      "DB00091\tCyclosporine\t111\t8.611493417346981\n",
      "DB00619\tImatinib\t108\t5.9956860686013975\n",
      "DB00773\tEtoposide\t102\t2.777712313988974\n",
      "DB00795\tSulfasalazine\t99\t3.315585808301857\n",
      "DB00806\tPentoxifylline\t99\t3.8725677755481107\n",
      "DB01645\tGenistein\t96\t4.153575578756933\n",
      "DB01611\tHydroxychloroquine\t90\t4.535518343138506\n",
      "DB01065\tMelatonin\t89\t3.5607650031360047\n",
      "DB00762\tIrinotecan\t88\t2.7665634463192\n",
      "DB00295\tMorphine\t87\t3.3667426007212593\n",
      "DB01137\tLevofloxacin\t87\t1.7565690698147964\n",
      "DB00741\tHydrocortisone\t87\t3.8173581511638077\n",
      "DB01233\tMetoclopramide\t84\t1.9707707099206266\n",
      "DB01183\tNaloxone\t83\t4.292251427403413\n",
      "DB01030\tTopotecan\t79\t2.114211397789907\n",
      "DB08910\tPomalidomide\t76\t1.676130543439113\n",
      "DB00959\tMethylprednisolone\t75\t2.6120392904094416\n",
      "DB00588\tFluticasone Propionate\t74\t2.4619681585018607\n",
      "DB00531\tCyclophosphamide\t72\t1.606216849292722\n",
      "DB00975\tDipyridamole\t72\t2.867203311238432\n",
      "DB00495\tZidovudine\t70\t3.819543546144744\n",
      "DB01200\tBromocriptine\t69\t3.2361200347446477\n",
      "DB00624\tTestosterone\t67\t2.26546332908827\n",
      "DB00688\tMycophenolate mofetil\t66\t1.2198019008352485\n",
      "DB00635\tPrednisone\t60\t1.7904146504671503\n",
      "DB01103\tQuinacrine\t59\t1.9281238111787005\n",
      "DB00993\tAzathioprine\t58\t1.2651128689977222\n",
      "DB01222\tBudesonide\t58\t2.967157793441156\n",
      "DB01042\tMelphalan\t58\t1.2453391840385615\n",
      "DB00694\tDaunorubicin\t58\t1.125798495565786\n",
      "DB00396\tProgesterone\t56\t1.225450121852841\n",
      "DB00917\tDinoprostone\t53\t2.6274748635094545\n",
      "DB00441\tGemcitabine\t52\t1.0248007492130935\n",
      "DB00958\tCarboplatin\t51\t1.1872548331725803\n",
      "DB01165\tOfloxacin\t48\t0.8364454646175277\n",
      "DB00608\tChloroquine\t47\t0.8858852272086288\n",
      "DB01118\tAmiodarone\t47\t0.8005396438713692\n",
      "DB00860\tPrednisolone\t44\t1.1015037059247128\n",
      "DB00305\tMitomycin\t44\t0.6699691864931889\n",
      "DB00908\tQuinidine\t43\t1.08873097429928\n",
      "DB01177\tIdarubicin\t38\t0.6468739601015208\n",
      "DB00502\tHaloperidol\t37\t1.7136023088309396\n",
      "DB00570\tVinblastine\t37\t0.8690853069826622\n",
      "DB00443\tBetamethasone\t35\t0.6710689256999115\n",
      "DB00977\tEthinyl Estradiol\t33\t0.5798858328791727\n",
      "DB01166\tCilostazol\t31\t0.5942939690819461\n",
      "DB04540\tCholesterol\t31\t4.37021954292773\n",
      "DB06774\tDB06774\t30\t0.8314755357939739\n",
      "DB01073\tFludarabine\t29\t0.6416084741930314\n",
      "DB01281\tAbatacept\t29\t1.8038215005188236\n",
      "DB01092\tOuabain\t29\t1.0853288111399086\n",
      "DB11672\tCurcumin\t28\t0.6031141154819399\n",
      "DB00904\tOndansetron\t28\t0.87229095563021\n",
      "DB00641\tSimvastatin\t27\t0.9988117773076428\n",
      "DB00242\tCladribine\t25\t0.48616605510832206\n",
      "DB01380\tCortisone acetate\t25\t0.37537049714630943\n",
      "DB01394\tColchicine\t25\t0.41966693868294974\n",
      "DB00445\tEpirubicin\t25\t0.3805217896944532\n",
      "DB00620\tTriamcinolone\t23\t0.42690189755132185\n",
      "DB01211\tClarithromycin\t23\t0.3755780729899096\n",
      "DB08880\tTeriflunomide\t23\t0.45212196813183086\n",
      "DB01024\tMycophenolic acid\t22\t0.4168803808182444\n",
      "DB01097\tLeflunomide\t22\t0.36941330127969757\n",
      "DB01039\tFenofibrate\t22\t1.0020189756553761\n",
      "DB00541\tVincristine\t22\t0.47043759541876895\n",
      "DB12116\tDB12116\t20\t0.6262304852282382\n",
      "DB00882\tClomifene\t20\t0.3813217738759036\n",
      "DB00822\tDisulfiram\t18\t0.3977862396511654\n",
      "DB01285\tCorticotropin\t18\t0.3252702684464074\n",
      "DB04630\tAldosterone\t18\t0.28511443369268397\n",
      "DB01015\tSulfamethoxazole\t17\t0.31089015508044704\n",
      "DB08904\tCertolizumab pegol\t16\t0.3849628049987272\n",
      "DB11526\tDB11526\t16\t0.5587032331154694\n",
      "DB00811\tRibavirin\t15\t0.5864270950229409\n",
      "DB00999\tHydrochlorothiazide\t15\t0.3196117771844759\n",
      "DB00108\tNatalizumab\t12\t0.9107381409373603\n",
      "DB00982\tIsotretinoin\t12\t0.19976199852124082\n",
      "DB01017\tMinocycline\t12\t0.1995703198940174\n",
      "DB04574\tEstropipate\t12\t0.33064855470228227\n",
      "DB00829\tDiazepam\t10\t1.2535547103321187\n",
      "DB00642\tPemetrexed\t9\t0.16278156330955715\n",
      "DB00286\tConjugated Estrogens\t9\t0.12282504608460838\n",
      "DB01045\tRifampicin\t9\t0.3761020223384842\n",
      "DB01241\tGemfibrozil\t9\t0.30270039573800966\n",
      "DB00158\tFolic Acid\t8\t0.10730235009968478\n",
      "DB00472\tFluoxetine\t8\t0.25490617192517884\n",
      "DB00591\tFluocinolone Acetonide\t8\t0.13452100767930436\n",
      "DB00640\tAdenosine\t7\t0.10278202474691567\n",
      "DB01656\tRoflumilast\t7\t0.109335123658115\n",
      "DB06273\tTocilizumab\t7\t0.1214646836186843\n",
      "DB01208\tSparfloxacin\t6\t0.07844716115244457\n",
      "DB06769\tDB06769\t6\t0.09218554819164575\n",
      "DB00074\tBasiliximab\t5\t0.41976292077098526\n",
      "DB01384\tParamethasone\t5\t0.09133599227408311\n",
      "DB09033\tDB09033\t5\t0.40329820452771276\n",
      "DB01068\tClonazepam\t4\t0.5349574676302318\n",
      "DB00552\tPentostatin\t4\t0.0611424149537204\n",
      "DB01167\tItraconazole\t4\t0.04939427873358444\n",
      "DB09341\tGlucose\t4\t0.07178607393753693\n",
      "DB04956\tAfelimomab\t4\t0.05977472982023033\n",
      "DB00470\tDronabinol\t3\t0.20531400966183572\n",
      "DB01223\tAminophylline\t3\t0.06429009654816106\n",
      "DB00290\tBleomycin\t3\t0.05143202709240445\n",
      "DB00764\tMometasone\t3\t0.05183982683982684\n",
      "DB01026\tKetoconazole\t3\t0.18214285714285713\n",
      "DB05459\tABT-874\t3\t0.13488045007032348\n",
      "DB05259\tGlatiramer Acetate\t3\t0.034366925064599485\n",
      "DB00105\tInterferon Alfa-2b, Recombinant\t2\t0.12439024390243902\n",
      "DB00008\tPeginterferon alfa-2a\t2\t0.19607843137254902\n",
      "DB00068\tInterferon beta-1b\t2\t0.3088235294117647\n",
      "DB00022\tPeginterferon alfa-2b\t2\t0.24285714285714285\n",
      "DB00033\tInterferon gamma-1b\t2\t1.1666666666666667\n",
      "DB14128\tNadide\t2\t1.2\n",
      "DB01410\tCiclesonide\t2\t0.02854251012145749\n",
      "DB01257\tEculizumab\t2\t0.030399274047186932\n",
      "DB02527\tCyclic Adenosine Monophosphate\t1\t0.011111111111111112\n",
      "DB01104\tSertraline\t1\t0.0196078431372549\n",
      "DB00143\tGlutathione\t1\t0.2\n",
      "DB00601\tLinezolid\t1\t0.01639344262295082\n",
      "DB00687\tFludrocortisone\t1\t0.016129032258064516\n",
      "DB01108\tTrilostane\t1\t0.05555555555555555\n",
      "DB11094\tVitamin D\t1\t0.06666666666666667\n",
      "DB00169\tCholecalciferol\t1\t0.14285714285714285\n",
      "DB00136\tCalcitriol\t1\t0.2\n"
     ]
    }
   ],
   "source": [
    "maxhit3=0\n",
    "drugs_in_top_k3={}\n",
    "drugsfr_in_top_k3={}\n",
    "for i in range(len(scores_per_gene3)):\n",
    "    score_g3=scores_per_gene3[i]\n",
    "    did_g3=dids_per_gene3[i]\n",
    "    idx_g3 = torch.flip(torch.argsort(score_g3), dims=[0]) #Ordena los índices de los elementos del tensor ordenados de mayor a menor\n",
    "    score_g3 = score_g3[idx_g3].numpy()\n",
    "    did_g3 = did_g3[idx_g3].numpy()\n",
    "    #print(did)\n",
    "    _, unique_indices3 = np.unique(did_g3, return_index=True) #Devuelve los índices donde aparecen por primera vez los valores únicos en did_g\n",
    "    topk=100\n",
    "    topk_indices3 = np.sort(unique_indices3)[:topk]  # Da los primeros 100 índices (o menos, si no hay 100) de los valores únicos en did_g, ordenados en forma ascendente.\n",
    "    proposed_did_g3 = did_g3[topk_indices3] #primeras 100 drogas\n",
    "    proposed_score_g3 = score_g3[topk_indices3] #primeros 100 scores\n",
    "    found_in_top_k3=0\n",
    "    found_drugs3=\"\\n\"\n",
    "    for j in range(topk):\n",
    "        drug_g3 = entity_id_map[int(proposed_did_g3[j])][10:17]\n",
    "        if drogas_ead_map.get(drug_g3, None) is not None:\n",
    "            found_in_top_k3+=1\n",
    "            score3 = proposed_score_g3[j]\n",
    "            if drug_g3 in drugs_in_top_k3:\n",
    "                drugs_in_top_k3[drug_g3]+=1\n",
    "                drugsfr_in_top_k3[drug_g3]+=1/(j+1)\n",
    "            else:\n",
    "                drugs_in_top_k3[drug_g3]=1\n",
    "                drugsfr_in_top_k3[drug_g3]=1/(j+1)\n",
    "            found_drugs3+=\"[{}]{}\\n\".format(j, drogas_ead_map[drug_g3])\n",
    "            #print(\"[{}]{}\".format(j, clinical_drug_map[drug]))\n",
    "    #print(\"{}\\t{}\".format(cov_related_genes[i], found_in_top_k))\n",
    "    if maxhit3< found_in_top_k3:\n",
    "        maxhit3=found_in_top_k3\n",
    "        maxgene3=genesEAD[i]\n",
    "        max_dugs3=found_drugs3\n",
    "print(\"{}\\t{}\\t{}\".format(maxgene3, maxhit3,max_dugs3)) #Se imprime el gen de los relacionados con las EADs con más hits de los medicamentos propuestos para dichas enfermedades,\n",
    "                                                     #el número de hits, y los nombres de los medicamentos asociados al gen.\n",
    "\n",
    "res3=[[drug, drogas_ead_map[drug] ,drugs_in_top_k3[drug],drugsfr_in_top_k3[drug]] for drug in drugs_in_top_k3.keys()]\n",
    "res3=reversed(sorted(res3, key=lambda x : x[2]))\n",
    "for drug in res3:\n",
    "    print(\"{}\\t{}\\t{}\\t{}\".format(drug[0], drug[1] ,drug[2],drug[3]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Wqrib5lHyI1W",
   "metadata": {
    "id": "Wqrib5lHyI1W"
   },
   "source": [
    "El gen con más hits, o relacionado con mayor cantidad dentro de los medicamentos analizados es el que posee ID Entrez 3439, o sea, IFNA1 ó interferon alpha 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "FovpVWPuOX6T",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1025,
     "status": "ok",
     "timestamp": 1729202493781,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "FovpVWPuOX6T",
    "outputId": "1a2a1006-b4e7-4983-f7d6-52dbc93104ec"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gene::3383\t61\t\n",
      "[0]Sunitinib\n",
      "[1]Cyclosporine\n",
      "[2]Doxorubicin\n",
      "[4]Sirolimus\n",
      "[6]Indomethacin\n",
      "[7]Cisplatin\n",
      "[8]Methotrexate\n",
      "[9]Simvastatin\n",
      "[10]Sulfasalazine\n",
      "[11]Thalidomide\n",
      "[13]Dasatinib\n",
      "[16]Cladribine\n",
      "[18]Tretinoin\n",
      "[19]Dexamethasone\n",
      "[21]Nilotinib\n",
      "[22]Docetaxel\n",
      "[25]Estradiol\n",
      "[26]Fludarabine\n",
      "[27]Imatinib\n",
      "[28]Paclitaxel\n",
      "[29]Dactinomycin\n",
      "[30]Tacrolimus\n",
      "[33]Prednisone\n",
      "[34]Topotecan\n",
      "[37]Mycophenolate mofetil\n",
      "[38]Pentoxifylline\n",
      "[39]Gemcitabine\n",
      "[40]Bleomycin\n",
      "[41]Budesonide\n",
      "[43]Etoposide\n",
      "[44]Dinoprostone\n",
      "[46]Fluticasone Propionate\n",
      "[47]Fenofibrate\n",
      "[49]Sulfamethoxazole\n",
      "[50]Carboplatin\n",
      "[51]Daunorubicin\n",
      "[52]Vinblastine\n",
      "[53]Hydroxychloroquine\n",
      "[56]Ribavirin\n",
      "[58]Epirubicin\n",
      "[61]Pazopanib\n",
      "[63]Melatonin\n",
      "[67]Mitomycin\n",
      "[68]Zidovudine\n",
      "[69]Mitoxantrone\n",
      "[70]Azathioprine\n",
      "[71]Methylprednisolone\n",
      "[73]Colchicine\n",
      "[75]Vincristine\n",
      "[78]Chloroquine\n",
      "[81]Adenosine\n",
      "[82]Morphine\n",
      "[84]Conjugated Estrogens\n",
      "[85]Minocycline\n",
      "[87]Quinacrine\n",
      "[88]Prednisolone\n",
      "[91]Ouabain\n",
      "[94]Cilostazol\n",
      "[95]Idarubicin\n",
      "[96]Pentostatin\n",
      "[97]Quercetin\n",
      "\n",
      "DB01268\tSunitinib\t181\t102.06854834848963\n",
      "DB01254\tDasatinib\t179\t18.30580525528253\n",
      "DB00997\tDoxorubicin\t179\t19.356622543419913\n",
      "DB01234\tDexamethasone\t176\t15.010913110945564\n",
      "DB00877\tSirolimus\t176\t9.445467748633465\n",
      "DB00091\tCyclosporine\t175\t35.87097215976116\n",
      "DB04868\tNilotinib\t174\t21.64157521899776\n",
      "DB00755\tTretinoin\t172\t12.490336192667149\n",
      "DB00783\tEstradiol\t171\t8.76568180790255\n",
      "DB06589\tPazopanib\t170\t11.582860852392844\n",
      "DB00515\tCisplatin\t161\t18.94403719775222\n",
      "DB00328\tIndomethacin\t157\t11.197263359769636\n",
      "DB01065\tMelatonin\t156\t9.715918200807483\n",
      "DB00619\tImatinib\t155\t9.350596968248334\n",
      "DB01041\tThalidomide\t152\t6.872303619014712\n",
      "DB00563\tMethotrexate\t152\t7.730532648023144\n",
      "DB00441\tGemcitabine\t145\t4.992707284382275\n",
      "DB01229\tPaclitaxel\t144\t5.701909368252499\n",
      "DB00445\tEpirubicin\t139\t3.6166626235346\n",
      "DB00242\tCladribine\t136\t6.173774564253601\n",
      "DB00570\tVinblastine\t122\t3.410132715175702\n",
      "DB01092\tOuabain\t116\t5.0977677227049645\n",
      "DB01248\tDocetaxel\t114\t3.787160682481808\n",
      "DB00958\tCarboplatin\t110\t3.3343547616344855\n",
      "DB00970\tDactinomycin\t108\t2.1769405908391515\n",
      "DB00773\tEtoposide\t107\t2.440322363315318\n",
      "DB00396\tProgesterone\t105\t2.2987375160112635\n",
      "DB00495\tZidovudine\t101\t3.517772256036941\n",
      "DB00688\tMycophenolate mofetil\t98\t1.8331766887960015\n",
      "DB00741\tHydrocortisone\t98\t4.024729591719923\n",
      "DB01030\tTopotecan\t96\t1.9724266903403254\n",
      "DB00917\tDinoprostone\t96\t6.019460014041371\n",
      "DB00864\tTacrolimus\t96\t1.9189080220114358\n",
      "DB00641\tSimvastatin\t90\t3.76944814736169\n",
      "DB01204\tMitoxantrone\t89\t1.8147471850569223\n",
      "DB01073\tFludarabine\t88\t3.187537174077701\n",
      "DB00795\tSulfasalazine\t85\t2.028422140069067\n",
      "DB00640\tAdenosine\t85\t1.6768658458079542\n",
      "DB00762\tIrinotecan\t84\t1.9662968597266026\n",
      "DB00694\tDaunorubicin\t84\t2.090732010097685\n",
      "DB01394\tColchicine\t83\t1.3902621365800394\n",
      "DB00811\tRibavirin\t82\t2.8457547800789498\n",
      "DB00290\tBleomycin\t80\t1.6491970588692537\n",
      "DB01137\tLevofloxacin\t79\t1.5313457035389597\n",
      "DB01017\tMinocycline\t79\t1.9133052667618913\n",
      "DB00588\tFluticasone Propionate\t78\t1.80370494570389\n",
      "DB01015\tSulfamethoxazole\t76\t1.5923437867665642\n",
      "DB00295\tMorphine\t76\t2.0773781820294617\n",
      "DB00635\tPrednisone\t75\t2.0101629766939433\n",
      "DB01611\tHydroxychloroquine\t74\t1.7671324864312572\n",
      "DB00982\tIsotretinoin\t72\t1.4965932323494782\n",
      "DB01103\tQuinacrine\t69\t1.9320384366862113\n",
      "DB01024\tMycophenolic acid\t67\t1.1012742886036473\n",
      "DB00624\tTestosterone\t66\t1.6511542504900716\n",
      "DB00541\tVincristine\t66\t1.4653419710382765\n",
      "DB01165\tOfloxacin\t63\t1.0051292799795373\n",
      "DB01177\tIdarubicin\t59\t1.0082273792702339\n",
      "DB00305\tMitomycin\t59\t0.9637725335332017\n",
      "DB00993\tAzathioprine\t58\t1.0562102293766342\n",
      "DB00286\tConjugated Estrogens\t56\t0.8976753069700457\n",
      "DB01222\tBudesonide\t54\t1.3580592442152533\n",
      "DB00136\tCalcitriol\t53\t1.0782503186247177\n",
      "DB01166\tCilostazol\t50\t0.8456655219386899\n",
      "DB00959\tMethylprednisolone\t49\t0.9378948567444736\n",
      "DB00999\tHydrochlorothiazide\t48\t1.0227918029625844\n",
      "DB01233\tMetoclopramide\t46\t0.8003313859869939\n",
      "DB00608\tChloroquine\t43\t0.6610446852148475\n",
      "DB00860\tPrednisolone\t43\t0.7445777047649738\n",
      "DB01183\tNaloxone\t41\t1.1193134213456566\n",
      "DB01039\tFenofibrate\t38\t1.285873839854703\n",
      "DB00443\tBetamethasone\t37\t0.6523419365085098\n",
      "DB00502\tHaloperidol\t37\t1.1263555597695734\n",
      "DB04216\tQuercetin\t33\t0.6647604459171783\n",
      "DB00975\tDipyridamole\t32\t0.8142000169324942\n",
      "DB00531\tCyclophosphamide\t31\t0.49583905355191965\n",
      "DB01118\tAmiodarone\t27\t0.3864308308377571\n",
      "DB01645\tGenistein\t26\t0.6538639929889788\n",
      "DB00143\tGlutathione\t24\t0.8792561597921648\n",
      "DB01200\tBromocriptine\t22\t1.1577071076844143\n",
      "DB08910\tPomalidomide\t22\t0.3451250700028901\n",
      "DB00806\tPentoxifylline\t21\t0.3088275569389991\n",
      "DB00908\tQuinidine\t21\t0.4021807828304602\n",
      "DB01042\tMelphalan\t21\t0.3239239690255692\n",
      "DB00472\tFluoxetine\t20\t0.4464317090492732\n",
      "DB00591\tFluocinolone Acetonide\t19\t0.2935709503814582\n",
      "DB00158\tFolic Acid\t18\t0.35533892701239894\n",
      "DB00620\tTriamcinolone\t15\t0.2468909437549762\n",
      "DB00169\tCholecalciferol\t15\t0.47786648796881936\n",
      "DB01656\tRoflumilast\t15\t0.1931845997596731\n",
      "DB00882\tClomifene\t14\t0.2638487840428865\n",
      "DB01097\tLeflunomide\t12\t0.1634334697501198\n",
      "DB00977\tEthinyl Estradiol\t12\t0.21265657230338847\n",
      "DB00074\tBasiliximab\t10\t0.32003249701400255\n",
      "DB00642\tPemetrexed\t10\t0.17280267472751576\n",
      "DB01410\tCiclesonide\t9\t0.11587930126856731\n",
      "DB00552\tPentostatin\t8\t0.11217451789150255\n",
      "DB00108\tNatalizumab\t8\t0.6580880099752109\n",
      "DB00904\tOndansetron\t7\t0.1409542478911529\n",
      "DB01241\tGemfibrozil\t7\t0.19259725444047476\n",
      "DB01167\tItraconazole\t7\t0.0907913695644765\n",
      "DB04540\tCholesterol\t7\t1.0040176732912092\n",
      "DB01281\tAbatacept\t6\t0.318596925166109\n",
      "DB01104\tSertraline\t6\t0.11239958701619199\n",
      "DB04574\tEstropipate\t6\t0.25210110559375265\n",
      "DB01211\tClarithromycin\t5\t0.05854651059027858\n",
      "DB00829\tDiazepam\t5\t1.0739944199023483\n",
      "DB00822\tDisulfiram\t5\t0.07758705116227912\n",
      "DB09341\tGlucose\t4\t0.059551452159811294\n",
      "DB04630\tAldosterone\t4\t0.04835308703796995\n",
      "DB01045\tRifampicin\t3\t0.0875222816399287\n",
      "DB11672\tCurcumin\t3\t0.040608546817733626\n",
      "DB00601\tLinezolid\t3\t0.0622836595846264\n",
      "DB01026\tKetoconazole\t3\t0.16520420902443372\n",
      "DB05459\tABT-874\t3\t0.1272934472934473\n",
      "DB12116\tDB12116\t3\t0.16787790697674418\n",
      "DB11526\tDB11526\t2\t0.042486583184257604\n",
      "DB01208\tSparfloxacin\t2\t0.027272727272727275\n",
      "DB00105\tInterferon Alfa-2b, Recombinant\t2\t0.07442348008385744\n",
      "DB00008\tPeginterferon alfa-2a\t2\t0.12040816326530612\n",
      "DB00033\tInterferon gamma-1b\t2\t1.0434782608695652\n",
      "DB00022\tPeginterferon alfa-2b\t2\t0.17045454545454547\n",
      "DB00068\tInterferon beta-1b\t2\t0.2976190476190476\n",
      "DB00470\tDronabinol\t2\t0.5333333333333333\n",
      "DB14128\tNadide\t2\t0.1607142857142857\n",
      "DB01384\tParamethasone\t2\t0.023088023088023088\n",
      "DB01380\tCortisone acetate\t2\t0.02548562548562549\n",
      "DB01068\tClonazepam\t1\t0.16666666666666666\n",
      "DB06774\tDB06774\t1\t0.011363636363636364\n",
      "DB00764\tMometasone\t1\t0.012987012987012988\n",
      "DB01223\tAminophylline\t1\t0.010526315789473684\n",
      "DB09033\tDB09033\t1\t0.03571428571428571\n",
      "DB00687\tFludrocortisone\t1\t0.013333333333333334\n",
      "DB11094\tVitamin D\t1\t0.021739130434782608\n",
      "DB01285\tCorticotropin\t1\t0.02631578947368421\n",
      "DB01108\tTrilostane\t1\t0.02702702702702703\n",
      "DB08904\tCertolizumab pegol\t1\t0.015151515151515152\n"
     ]
    }
   ],
   "source": [
    "maxhit4=0\n",
    "drugs_in_top_k4={}\n",
    "drugsfr_in_top_k4={}\n",
    "for i in range(len(scores_per_gene4)):\n",
    "    score_g4=scores_per_gene4[i]\n",
    "    did_g4=dids_per_gene4[i]\n",
    "    idx_g4 = torch.flip(torch.argsort(score_g4), dims=[0]) #Ordena los índices de los elementos del tensor ordenados de mayor a menor\n",
    "    score_g4 = score_g4[idx_g4].numpy()\n",
    "    did_g4 = did_g4[idx_g4].numpy()\n",
    "    #print(did)\n",
    "    _, unique_indices4 = np.unique(did_g4, return_index=True) #Devuelve los índices donde aparecen por primera vez los valores únicos en did_g\n",
    "    topk=100\n",
    "    topk_indices4 = np.sort(unique_indices4)[:topk]  # Da los primeros 100 índices (o menos, si no hay 100) de los valores únicos en did_g, ordenados en forma ascendente.\n",
    "    proposed_did_g4 = did_g4[topk_indices4] #primeras 100 drogas\n",
    "    proposed_score_g4 = score_g4[topk_indices4] #primeros 100 scores\n",
    "    found_in_top_k4=0\n",
    "    found_drugs4=\"\\n\"\n",
    "    for j in range(topk):\n",
    "        drug_g4 = entity_id_map[int(proposed_did_g4[j])][10:17]\n",
    "        if drogas_ead_map.get(drug_g4, None) is not None:\n",
    "            found_in_top_k4+=1\n",
    "            score4 = proposed_score_g4[j]\n",
    "            if drug_g4 in drugs_in_top_k4:\n",
    "                drugs_in_top_k4[drug_g4]+=1\n",
    "                drugsfr_in_top_k4[drug_g4]+=1/(j+1)\n",
    "            else:\n",
    "                drugs_in_top_k4[drug_g4]=1\n",
    "                drugsfr_in_top_k4[drug_g4]=1/(j+1)\n",
    "            found_drugs4+=\"[{}]{}\\n\".format(j, drogas_ead_map[drug_g4])\n",
    "            #print(\"[{}]{}\".format(j, clinical_drug_map[drug]))\n",
    "    #print(\"{}\\t{}\".format(cov_related_genes[i], found_in_top_k))\n",
    "    if maxhit4< found_in_top_k4:\n",
    "        maxhit4=found_in_top_k4\n",
    "        maxgene4=genesEAD[i]\n",
    "        max_dugs4=found_drugs4\n",
    "print(\"{}\\t{}\\t{}\".format(maxgene4, maxhit4,max_dugs4)) #Se imprime el gen de los relacionados con las EADs con más hits de los medicamentos propuestos para dichas enfermedades,\n",
    "                                                     #el número de hits, y los nombres de los medicamentos asociados al gen.\n",
    "\n",
    "res4=[[drug, drogas_ead_map[drug] ,drugs_in_top_k4[drug],drugsfr_in_top_k4[drug]] for drug in drugs_in_top_k4.keys()]\n",
    "res4=reversed(sorted(res4, key=lambda x : x[2]))\n",
    "for drug in res4:\n",
    "    print(\"{}\\t{}\\t{}\\t{}\".format(drug[0], drug[1] ,drug[2],drug[3]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "H5shERHjyVsX",
   "metadata": {
    "id": "H5shERHjyVsX"
   },
   "source": [
    "El gen con más hits, o relacionado con mayor cantidad dentro de los medicamentos analizados es el que posee ID Entrez 3383, o sea, ICAM1 ó intercellular adhesion molecule 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "QPYTXxpC2Y0h",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1859,
     "status": "ok",
     "timestamp": 1729202499921,
     "user": {
      "displayName": "Nicolás Barrera",
      "userId": "16526758339075686419"
     },
     "user_tz": 180
    },
    "id": "QPYTXxpC2Y0h",
    "outputId": "21fd2a71-149d-428e-de1b-71330f77c506"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gene::3439\t73\t\n",
      "[0]Tretinoin\n",
      "[1]Thalidomide\n",
      "[2]Methotrexate\n",
      "[3]Dexamethasone\n",
      "[4]Zidovudine\n",
      "[5]Ribavirin\n",
      "[6]Paclitaxel\n",
      "[7]Cisplatin\n",
      "[8]Gemcitabine\n",
      "[9]Doxorubicin\n",
      "[10]Hydrocortisone\n",
      "[11]Sirolimus\n",
      "[12]Prednisone\n",
      "[13]Fludarabine\n",
      "[14]Methylprednisolone\n",
      "[15]Cyclosporine\n",
      "[16]Chloroquine\n",
      "[17]Etoposide\n",
      "[18]Topotecan\n",
      "[19]Prednisolone\n",
      "[20]Vincristine\n",
      "[21]Isotretinoin\n",
      "[22]Azathioprine\n",
      "[23]Curcumin\n",
      "[24]Dinoprostone\n",
      "[25]Hydroxychloroquine\n",
      "[27]Carboplatin\n",
      "[28]Budesonide\n",
      "[29]Imatinib\n",
      "[30]Colchicine\n",
      "[31]Estradiol\n",
      "[32]Mitoxantrone\n",
      "[33]Dactinomycin\n",
      "[35]Sunitinib\n",
      "[36]Quercetin\n",
      "[37]Testosterone\n",
      "[40]Cyclophosphamide\n",
      "[41]Pazopanib\n",
      "[43]Dasatinib\n",
      "[44]Genistein\n",
      "[45]Sulfamethoxazole\n",
      "[46]Indomethacin\n",
      "[47]Melatonin\n",
      "[48]DB12116\n",
      "[49]Pentoxifylline\n",
      "[50]Cladribine\n",
      "[51]Mycophenolate mofetil\n",
      "[54]Vinblastine\n",
      "[55]Simvastatin\n",
      "[56]Mitomycin\n",
      "[57]Sulfasalazine\n",
      "[58]Idarubicin\n",
      "[59]Roflumilast\n",
      "[64]Bleomycin\n",
      "[65]Vitamin D\n",
      "[66]Morphine\n",
      "[67]Progesterone\n",
      "[68]Tacrolimus\n",
      "[69]Irinotecan\n",
      "[72]Epirubicin\n",
      "[74]Daunorubicin\n",
      "[76]Fenofibrate\n",
      "[81]Nilotinib\n",
      "[83]Cyclic Adenosine Monophosphate\n",
      "[85]Adenosine\n",
      "[86]Mycophenolic acid\n",
      "[88]Fluticasone Propionate\n",
      "[89]Itraconazole\n",
      "[90]Diazepam\n",
      "[91]Betamethasone\n",
      "[92]Cholesterol\n",
      "[96]Naloxone\n",
      "[97]Metoclopramide\n",
      "\n",
      "DB01234\tDexamethasone\t692\t87.32312936729433\n",
      "DB00755\tTretinoin\t671\t69.44752768183011\n",
      "DB00783\tEstradiol\t631\t31.553799340585766\n",
      "DB00877\tSirolimus\t625\t36.88553964371083\n",
      "DB00997\tDoxorubicin\t619\t41.66507715937413\n",
      "DB01229\tPaclitaxel\t609\t48.21567378811774\n",
      "DB00515\tCisplatin\t593\t62.20284233104136\n",
      "DB01268\tSunitinib\t579\t144.44445093134945\n",
      "DB06589\tPazopanib\t575\t27.13747742433817\n",
      "DB00563\tMethotrexate\t556\t38.57507622984159\n",
      "DB00091\tCyclosporine\t547\t70.0214721033484\n",
      "DB00328\tIndomethacin\t538\t34.617417906864006\n",
      "DB01254\tDasatinib\t521\t33.10223355598196\n",
      "DB01065\tMelatonin\t516\t30.84106056379132\n",
      "DB01041\tThalidomide\t504\t23.98041444127062\n",
      "DB04216\tQuercetin\t491\t41.678218826988605\n",
      "DB04868\tNilotinib\t465\t37.87456781127466\n",
      "DB00970\tDactinomycin\t421\t10.309391839425581\n",
      "DB00441\tGemcitabine\t393\t15.996482708486534\n",
      "DB00741\tHydrocortisone\t382\t21.305269358774677\n",
      "DB00619\tImatinib\t378\t20.474976111522185\n",
      "DB01645\tGenistein\t378\t22.400585770303064\n",
      "DB00773\tEtoposide\t377\t8.99713074974172\n",
      "DB00795\tSulfasalazine\t373\t10.725503458356457\n",
      "DB00396\tProgesterone\t342\t9.289239837034065\n",
      "DB01248\tDocetaxel\t338\t12.305640138557951\n",
      "DB01204\tMitoxantrone\t322\t9.549175626642711\n",
      "DB00958\tCarboplatin\t322\t8.85098934488775\n",
      "DB00917\tDinoprostone\t318\t23.014581894237704\n",
      "DB00641\tSimvastatin\t300\t15.527552620627324\n",
      "DB11672\tCurcumin\t298\t27.973925473685345\n",
      "DB00635\tPrednisone\t295\t9.659569611814197\n",
      "DB00570\tVinblastine\t293\t7.996717086868276\n",
      "DB00495\tZidovudine\t290\t11.466521377349377\n",
      "DB00624\tTestosterone\t289\t7.983373212456034\n",
      "DB00959\tMethylprednisolone\t287\t9.862953443798261\n",
      "DB00864\tTacrolimus\t280\t6.83970548315175\n",
      "DB00688\tMycophenolate mofetil\t279\t4.992204649340326\n",
      "DB01030\tTopotecan\t273\t5.974245330213222\n",
      "DB01394\tColchicine\t260\t4.666918493043051\n",
      "DB01222\tBudesonide\t258\t10.856422057356266\n",
      "DB00806\tPentoxifylline\t257\t8.03924622283301\n",
      "DB01092\tOuabain\t257\t10.169220265809365\n",
      "DB00242\tCladribine\t248\t9.021621123923916\n",
      "DB01611\tHydroxychloroquine\t247\t7.87034412477278\n",
      "DB00295\tMorphine\t246\t7.581997248686092\n",
      "DB00993\tAzathioprine\t243\t4.886050715757258\n",
      "DB00640\tAdenosine\t236\t8.7845214568239\n",
      "DB00694\tDaunorubicin\t235\t4.9226718049973615\n",
      "DB04540\tCholesterol\t235\t31.739707062550163\n",
      "DB00762\tIrinotecan\t233\t6.142640388654911\n",
      "DB06774\tDB06774\t223\t7.422283227592341\n",
      "DB00860\tPrednisolone\t220\t6.140664634751136\n",
      "DB00305\tMitomycin\t220\t4.2769105022868645\n",
      "DB01073\tFludarabine\t217\t6.536846547160125\n",
      "DB00608\tChloroquine\t217\t4.421336284060783\n",
      "DB00982\tIsotretinoin\t206\t4.7895795398222045\n",
      "DB00445\tEpirubicin\t205\t4.622279156781135\n",
      "DB00975\tDipyridamole\t205\t6.119032463043355\n",
      "DB00541\tVincristine\t205\t4.526417867876112\n",
      "DB01103\tQuinacrine\t201\t5.440981997697607\n",
      "DB01183\tNaloxone\t198\t7.777964142747851\n",
      "DB01017\tMinocycline\t188\t4.611932841636527\n",
      "DB09341\tGlucose\t187\t10.476894385700618\n",
      "DB12116\tDB12116\t186\t28.62232258208239\n",
      "DB00811\tRibavirin\t185\t6.898469114748601\n",
      "DB11094\tVitamin D\t184\t10.028502833542069\n",
      "DB01233\tMetoclopramide\t183\t3.6749590126711373\n",
      "DB00588\tFluticasone Propionate\t181\t4.860563194898254\n",
      "DB01137\tLevofloxacin\t176\t3.49226893173154\n",
      "DB08910\tPomalidomide\t173\t3.494926415284546\n",
      "DB01039\tFenofibrate\t171\t5.678434476989911\n",
      "DB02527\tCyclic Adenosine Monophosphate\t157\t9.531283880242645\n",
      "DB01118\tAmiodarone\t156\t2.393488764958395\n",
      "DB04630\tAldosterone\t154\t2.8724016401350503\n",
      "DB00908\tQuinidine\t143\t3.1750920289213034\n",
      "DB01015\tSulfamethoxazole\t142\t2.7968743541991685\n",
      "DB01165\tOfloxacin\t136\t2.19153407764483\n",
      "DB00502\tHaloperidol\t131\t4.888022493234107\n",
      "DB00286\tConjugated Estrogens\t130\t2.3333389685564083\n",
      "DB00999\tHydrochlorothiazide\t125\t2.561537352398448\n",
      "DB01177\tIdarubicin\t124\t2.0705086476786327\n",
      "DB01200\tBromocriptine\t121\t5.68503477428653\n",
      "DB01166\tCilostazol\t120\t2.1549707296988805\n",
      "DB00531\tCyclophosphamide\t116\t2.33445079401181\n",
      "DB01024\tMycophenolic acid\t114\t1.9146513944057013\n",
      "DB04115\tBerberine\t114\t5.817479453259252\n",
      "DB00443\tBetamethasone\t113\t2.0040713611160363\n",
      "DB00136\tCalcitriol\t109\t2.4801343288319218\n",
      "DB04574\tEstropipate\t107\t2.8642183461804662\n",
      "DB00143\tGlutathione\t107\t4.094319139452919\n",
      "DB00977\tEthinyl Estradiol\t105\t1.8209540207951447\n",
      "DB00158\tFolic Acid\t102\t2.3722100642305617\n",
      "DB00290\tBleomycin\t99\t2.0611007181553247\n",
      "DB08904\tCertolizumab pegol\t96\t2.6510608233505617\n",
      "DB01042\tMelphalan\t91\t1.7479978105981069\n",
      "DB01211\tClarithromycin\t76\t1.294003668343883\n",
      "DB00620\tTriamcinolone\t74\t1.3118184234297923\n",
      "DB01656\tRoflumilast\t72\t1.069182051384102\n",
      "DB00882\tClomifene\t62\t1.1066292422292965\n",
      "DB00472\tFluoxetine\t56\t1.4550029364409698\n",
      "DB00591\tFluocinolone Acetonide\t56\t0.8767963124412735\n",
      "DB00822\tDisulfiram\t53\t1.0349265024538523\n",
      "DB08880\tTeriflunomide\t52\t0.9320940437162892\n",
      "DB00904\tOndansetron\t48\t1.287239924918037\n",
      "DB01281\tAbatacept\t48\t3.0822821709657875\n",
      "DB01097\tLeflunomide\t43\t0.6513396160317361\n",
      "DB01045\tRifampicin\t43\t1.1834515783215973\n",
      "DB00829\tDiazepam\t41\t4.738983484758785\n",
      "DB01167\tItraconazole\t38\t0.4921815268423964\n",
      "DB00169\tCholecalciferol\t38\t1.6735959184446982\n",
      "DB01241\tGemfibrozil\t38\t1.0382878910754039\n",
      "DB06273\tTocilizumab\t36\t0.7087603186389917\n",
      "DB01380\tCortisone acetate\t32\t0.4773698554658926\n",
      "DB00108\tNatalizumab\t31\t2.0472033020818974\n",
      "DB14128\tNadide\t31\t6.010519320549608\n",
      "DB00074\tBasiliximab\t30\t1.608855821426717\n",
      "DB00642\tPemetrexed\t30\t0.5167391652642265\n",
      "DB11526\tDB11526\t29\t1.059716675949348\n",
      "DB01285\tCorticotropin\t23\t0.4390957894292362\n",
      "DB14104\tLinoleic acid\t20\t0.3333116659938735\n",
      "DB00470\tDronabinol\t19\t1.813894968820017\n",
      "DB06769\tDB06769\t18\t0.36082481280674\n",
      "DB00552\tPentostatin\t17\t0.2398002900594534\n",
      "DB01410\tCiclesonide\t16\t0.2202876105192782\n",
      "DB01104\tSertraline\t16\t0.29414729978919874\n",
      "DB01026\tKetoconazole\t14\t0.5832362325797757\n",
      "DB04956\tAfelimomab\t13\t0.18885598883911203\n",
      "DB09033\tDB09033\t13\t0.7417653640029368\n",
      "DB05459\tABT-874\t12\t0.7202729546705744\n",
      "DB01384\tParamethasone\t10\t0.16652811281504087\n",
      "DB00068\tInterferon beta-1b\t9\t1.4395409479581307\n",
      "DB00022\tPeginterferon alfa-2b\t9\t2.1323287823287824\n",
      "DB01208\tSparfloxacin\t8\t0.10571988842517185\n",
      "DB01068\tClonazepam\t8\t1.546068578741343\n",
      "DB00105\tInterferon Alfa-2b, Recombinant\t7\t0.4087483645091723\n",
      "DB00008\tPeginterferon alfa-2a\t7\t0.6357848402518902\n",
      "DB00033\tInterferon gamma-1b\t7\t3.5434782608695654\n",
      "DB05259\tGlatiramer Acetate\t6\t0.06674435337172252\n",
      "DB00764\tMometasone\t5\t0.07852546996382613\n",
      "DB01223\tAminophylline\t5\t0.09068942821065062\n",
      "DB00601\tLinezolid\t5\t0.09183499694441932\n",
      "DB01108\tTrilostane\t4\t0.1288983720562668\n",
      "DB00687\tFludrocortisone\t2\t0.02946236559139785\n",
      "DB01257\tEculizumab\t2\t0.030399274047186932\n",
      "DB14783\tDiroximel fumarate\t2\t0.05505952380952381\n",
      "DB11988\tDB11988\t1\t0.010101010101010102\n",
      "DB09122\tPeginterferon beta-1a\t1\t0.010638297872340425\n",
      "DB13038\tGlyceryl trioleate\t1\t0.01694915254237288\n"
     ]
    }
   ],
   "source": [
    "maxhit5=0\n",
    "drugs_in_top_k5={}\n",
    "drugsfr_in_top_k5={}\n",
    "for i in range(len(scores_per_gene5)):\n",
    "    score_g5=scores_per_gene5[i]\n",
    "    did_g5 = dids_per_gene5[i]\n",
    "    idx_g5 = torch.flip(torch.argsort(score_g5), dims=[0]) #Ordena los índices de los elementos del tensor ordenados de mayor a menor\n",
    "    score_g5 = score_g5[idx_g5].numpy()\n",
    "    did_g5 = did_g5[idx_g5].numpy()\n",
    "    #print(did)\n",
    "    _, unique_indices5 = np.unique(did_g5, return_index=True) #Devuelve los índices donde aparecen por primera vez los valores únicos en did_g\n",
    "    topk=100\n",
    "    topk_indices5 = np.sort(unique_indices5)[:topk]  # Da los primeros 100 índices (o menos, si no hay 100) de los valores únicos en did_g, ordenados en forma ascendente.\n",
    "    proposed_did_g5 = did_g5[topk_indices5] #primeras 100 drogas\n",
    "    proposed_score_g5 = score_g5[topk_indices5] #primeros 100 scores\n",
    "    found_in_top_k5=0\n",
    "    found_drugs5=\"\\n\"\n",
    "    for j in range(topk):\n",
    "        drug_g5 = entity_id_map[int(proposed_did_g5[j])][10:17]\n",
    "        if drogas_ead_map.get(drug_g5, None) is not None:\n",
    "            found_in_top_k5+=1\n",
    "            score5 = proposed_score_g5[j]\n",
    "            if drug_g5 in drugs_in_top_k5:\n",
    "                drugs_in_top_k5[drug_g5]+=1\n",
    "                drugsfr_in_top_k5[drug_g5]+=1/(j+1)\n",
    "            else:\n",
    "                drugs_in_top_k5[drug_g5]=1\n",
    "                drugsfr_in_top_k5[drug_g5]=1/(j+1)\n",
    "            found_drugs5+=\"[{}]{}\\n\".format(j, drogas_ead_map[drug_g5])\n",
    "            #print(\"[{}]{}\".format(j, clinical_drug_map[drug]))\n",
    "    #print(\"{}\\t{}\".format(cov_related_genes[i], found_in_top_k))\n",
    "    if maxhit5< found_in_top_k5:\n",
    "        maxhit5=found_in_top_k5\n",
    "        maxgene5=genesEAD[i]\n",
    "        max_dugs5=found_drugs5\n",
    "print(\"{}\\t{}\\t{}\".format(maxgene5, maxhit5,max_dugs5)) #Se imprime el gen de los relacionados con las EADs con más hits de los medicamentos propuestos para dichas enfermedades,\n",
    "                                                     #el número de hits, y los nombres de los medicamentos asociados al gen.\n",
    "\n",
    "res5=[[drug, drogas_ead_map[drug] ,drugs_in_top_k5[drug],drugsfr_in_top_k5[drug]] for drug in drugs_in_top_k5.keys()]\n",
    "res5=reversed(sorted(res5, key=lambda x : x[2]))\n",
    "for drug in res5:\n",
    "    print(\"{}\\t{}\\t{}\\t{}\".format(drug[0], drug[1] ,drug[2],drug[3]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "uSSZKkv-4yXO",
   "metadata": {
    "id": "uSSZKkv-4yXO"
   },
   "source": [
    "El gen con más hits, o relacionado con mayor cantidad dentro de los medicamentos analizados es el que posee ID Entrez 3439, o sea, IFNA1 ó interferon alpha 1."
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "gpuType": "L4",
   "machine_shape": "hm",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}