GOenrichment

GOenrichments is a package that facilitates GO set enrichment analysis and subsequent network visualization. The significance of the GO enrichments are estimated using the hypergeometric test based on the query set of genes passing the user-defined fitness score threshold compared to the genes in the background set (the gene universe). Because of the high degree of redundancy between GO terms, an advantageous of the GOenrichment package is to simplify the output of a GO enrichment analysis. Specifically, clusters in the enrichment network are defned by GO terms that overlap by >= 50%. Node size is proportional to the significane of the GO enrichment (FDR score) and edge weight are proportional to the degree of overlap between the GO terms. In each cluster we highlight the most significantly enriched GO term by increasing the fontsize and setting the fontface to bold. In this manner, cellular and functional modules emerge in the network, defined by tightly connected color coordinated clusters connected by multiple edges and a holistic view of the cellular response to the perturbation ofo interest is possible.

Installation

You can install the development version of GOenrichment like so:

devtools::install_github("gurinina/GOenrichment")

Example

This is a basic example which shows you how to solve a common problem:

library(GOenrichment)

this is a GO enrichment for a yeast fitness dataset, run in presence of a drug called azithromycin

runGORESP produces the enrichments

goresp = runGORESP(mat=sampleFitdata,coln=1,curr_exp = colnames(sampleFitdata)[1],sig = 1,fdrThresh = 0.2,bp_path = NULL,bp_input = yGOBP.gmt,go_path = NULL,go_input = dfGOBP)

visSetup prepares the output of runGORESP for downstream plotting

vis = visSetup(goresp$enrichInfo,goresp$edgeMat```

runNetwork produces the GO enrichment network

runNetwork(vis$nodes,vis$edges)