Merge branch 'auto-sample-sheet' into 'dev'

Auto sample sheet See merge request epi2melabs/workflow-containers/wf-artic!25
epi2me-labs · May 10, 2021 · 5d4a782 · 5d4a782
2 parents f7dba00 + 49629c3
commit 5d4a782
Show file tree

Hide file tree

Showing 3 changed files with 54 additions and 5 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -9,6 +9,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Ability to configure depth coverage reporting value.
 - Add explicit pins of conda packages.
 - Inclusion of SpikeSeq workflow, and reporting.
+- Optional auto-detection of sample_id
 
 ## [v0.1.4]
 ### Changed

diff --git a/main.nf b/main.nf
@@ -1,4 +1,5 @@
 #!/usr/bin/env nextflow
+import java.util.zip.GZIPInputStream;
 
 nextflow.enable.dsl = 2
 
@@ -31,6 +32,8 @@ Options:
                                   (e.g. 89% genome covered at > `report_depth`)
                                   indicating correspondence between
     --genotype_variants   FILE    A VCF containing known variants to genotype.
+    --detect_samples      BOOL    Automatically determine sample_id information from fastq
+                                  header, replaces the --samples csv.
     --report_clade        BOOL    Show results of Nextclade analysis in report.
     --report_lineage      BOOL    Show results of Pangolin analysis in report.
     --report_coverage     BOOL    Show genome coverage traces in report.
@@ -53,6 +56,29 @@ Notes:
 }
 
 
+def sampleFromHeader(fastq) {
+    InputStream stream
+    if (fastq.endsWith("gz")) {
+        stream = new GZIPInputStream(
+            new FileInputStream(fastq.toString()));
+    } else {
+        stream = new FileInputStream(fastq.toString());
+    }
+
+    Reader reader = new InputStreamReader(stream, "UTF-8");
+    def line = reader.readLine()
+    def found = line.findAll("sample_id=[A-Za-z0-9]+")
+    reader.close()
+
+    if (found) {
+       return found[0].split('=')[1]
+    }
+
+    println("--detect_samples is on but no sample_id was found in fastq header")
+    exit 1
+}
+
+
 process checkSampleSheet {
     label "artic"
     cpus 1
@@ -337,6 +363,11 @@ workflow {
         exit 1
     }
 
+    if (params.samples && params.detect_samples) {
+        println("Select either `--samples` or `--detect_samples`, not both")
+        exit 1
+    }
+
     if (!params.min_len) {
         params.remove('min_len')
         if (params.scheme_version == "V1200") {
@@ -374,6 +405,7 @@ workflow {
     primers = file(
         "${scheme_directory}/${params.full_scheme_name}/${params.scheme_name}.scheme.bed",
         type:'file', checkIfExists:true)
+
     // check sample sheet
     sample_sheet = null
     if (params.samples) {
@@ -385,6 +417,7 @@ workflow {
 
     // resolve whether we have demultiplexed data or single sample
     not_barcoded = file("$params.fastq/*.fastq*", type: 'file', maxdepth: 1)
+
     // remove empty barcode_dirs
     barcode_dirs = file("$params.fastq/barcode*", type: 'dir', maxdepth: 1)
     valid_barcode_dirs = []
@@ -396,6 +429,7 @@ workflow {
             valid_barcode_dirs << d
         }
     }
+
     if (barcode_dirs) {
         println("Found barcode directories")
         if (invalid_barcode_dirs.size() > 0) {
@@ -405,11 +439,24 @@ workflow {
             }
         }
         if (!sample_sheet) {
-            // just map directory name to self
-            sample_sheet = Channel
-                .fromPath(valid_barcode_dirs)
-                .filter(~/.*barcode[0-9]{1,3}$/)  // up to 192
-                .map { path -> tuple(path.baseName, path.baseName) }
+            // check if we need to autodetect sample names
+            if (params.detect_samples) {
+                sample_sheet = Channel
+                    .fromPath(valid_barcode_dirs)
+                    .map { path -> tuple(
+                        path.baseName, 
+                        file("${path}{**,.}/*.{fastq,fastq.gz,fq,fq.gz}")) }
+                    .filter { pathTuple -> !pathTuple[1].isEmpty() }
+                    .map { pathTuple -> tuple(
+                        pathTuple[0],
+                        sampleFromHeader(pathTuple[1][0])) }
+            } else {
+                // just map directory name to self
+                sample_sheet = Channel
+                    .fromPath(valid_barcode_dirs)
+                    .filter(~/.*barcode[0-9]{1,3}$/)  // up to 192
+                    .map { path -> tuple(path.baseName, path.baseName) }
+            }
         }
         Channel
             .fromPath(valid_barcode_dirs)

diff --git a/nextflow.config b/nextflow.config
@@ -21,6 +21,7 @@ params {
     scheme_name = "SARS-CoV-2"
     scheme_version = "V3"
     genotype_variants = null
+    detect_samples = false
     report_clade = true
     report_lineage = true
     report_coverage = true