MSGMM

R software scripts for fitting Gaussian mixture models to multiple data samples simultaneously, using the EM algorithm. It makes use of C++ code via the R package Rcpp. MSGMM also takes advantage of the open source C++ linear algebra library 'Armadillo' via the R package RcppArmadillo. MSGMM also uses the data.table package. These packages need to be installed first in order to use MSGMM.

The method implemented in MSGMM is detailed in: Computationally efficient multi-sample flow cytometry data analysis using Gaussian mixture models

Philip Rutten, Tim Robert Mocking, Jacqueline Cloos, Wessel van Wieringen, Costa Bachas

Running the code

MSGMM is not an R package. Store the msgmm.R file in the same directory as your R script, or specify the full path name. Source the R script:

> source("msgmm.R")

The msgmm.R file in turn includes C++ code from external C++ files in a similar fashion using the sourceCpp("...") function. Store the C++ files called inside sourceCpp() in the same directory as the msgmm.R script, or specify the full path name.

See a demonstration of MSGMM analysis on simulated sample files in the demo_analysis.R script.

The main function fitMSGMM() fits a GMM to multiple samples, using the EM algorithm, and returns estimated model parameters. fitMSGMM iterates data files multiple times and at each iteration opens and closes them sequentially. Files should be preprocessed and in CSV format. See the reference manual for detailed usage information.