OmicScan

OmicScan is an R package designed for biomarker discovery from RNA-seq data. The pipeline processes gene count data, performs a multi-step process for gene filtering, and finally performs biomarker discovery using a suite of machine learning algorithms.

Installation

To install the package from GitHub, use:

# Install devtools if not already installed
install.packages("devtools")

# Install OmicScan from GitHub
devtools::install_github("AwanyDenis/OmicScan")

# Once installed, load it via
library(OmicScan)

Running

To run the software, at least two inputs are required: a gene expression data in raw count (non-normalized) form, and a metadata containing the groups (e.g. case/control) for the samples.

help(OmicScan)

res <- OmicScan(expMat = gene_counts, metaData = metaData, sampleID = 'PID',phenotype = 'Phenotype',cohort = NULL, 
     contr = 'Control', trainingSize=0.7, batchCorrect = FALSE,
     covariate = NULL, normalizeData  = TRUE, 
     normalizeMethod = 'log2-cpm', filterExprMat = FALSE, 
     trainModel = 'SVM', shapley.ratio = 0.5, permut.import.ratio = 0.5)

print(res$final.signature)
print(res$performance.metric)