rPEGEOS: Pathway Enrichment for Gene ExpressiOn Signatures with R

This package performs gene set enrichment for a given gene expression signature based on the TF-IDF matrix, computed on the pathway representations. rPEGEOS comes pre-packaged with 16 different gene sets on human gene sets. Other species can be accomodated. Contact me and we'll chat.

Installing

In R, do:

library(devtools)
devtools::install_github("diogocamacho/rpegeos")

Running rPEGEOS

The easiest way to run rPEGEOS is through its wrapper enrich_geneset. All you need is a set of genes that you want to enrich, as Entrez gene IDs.

Wrapper

To make it easier to run, a wrapper is also deployed. To enrich a gene set (provided you have a tf-idf matrix and a cross-product vector, do:

ES <- enrich_geneset(gene_set)

where gene_set is an Nx3 matrix where the columns are:

• Column 1: Entrez IDs
• Column 2: Log2(fold change)
• Column 3: p-value (or adjusted p-value) for differential expression

There is no need to pre-filter the gene set, as rPEGEOS will do that for you, based on thresholds defined by the user. Follow on-screen instructions and you are all set!

Plotting et al.

The protocol above will generate all the necessary elements to generate a final results data frame. The results can be plotted with ggplot2 or your favorite plotting tools. A plot of the top enrichment results will be generated as a consequence of running the wrapper.

In the works

An R Shiny package is being developed to perform gene set enrichments as a web service.