bic.run.gsa: Run gene set analysis using the package 'piano'

In bic-mskcc/bicrnaseq: MSKCC Bioinformatics Core RNASeq

Description

Run gene set analysis for one pairwise comparison of sample groups. GSA is run using the R package 'piano'. For each gene set, means are calculated for the genes with positive fold changes and for genes with negative fold changes. P-values and adjusted p-values are then calculated and gene sets with a significant number of of up- or down-regulated genes are reported in the output. Gene Sets included are from the MSigDB Collections:

Usage

bic.run.gsa(species, deseq.res, min.gns = 5, max.gns = 1000, max.p = 0.1,
  nPerm = 10000, fcQ = T, fc2keep = log2(1.5), frac2keep = 4)

Arguments

species	currently only human and mouse are supported
deseq.res	the matrix of ALL DESeq results (unfiltered) in BIC format (GeneID,GeneSymbol,pval,P.adj, log2[condB/condA],Mean_at_cond_condA, Mean_at_cond_condB
min.gns	minimum number of genes a set must have in order to be included in analysis; Default: 5
max.gns	maximum number of genes a set must have in order to be included in analysis; Default: 1000
max.p	maximum p-value to use as cut off; Default: 0.1
nPerm	the number of times the genes are randomized; Default: 1e4
fcQ	filter gene sets based on fc2keep and frac2keep Default: TRUE
fc2keep	fold change cutoff (see frac2keep); Default: log2(1.5)
frac2keep	at least 1/frac2keep of genes in a set must have fold change >= abs(fc2keep); Default: 4

Details

C1: positional gene sets C2: curated gene sets C3: motif gene sets C5: GO gene sets C6: oncogenic signatures C7: immunologic signatures

Note: Mouse gene sets were downloaded from http://bioinf.wehi.edu.au/software/MSigDB/

Value

a list of two matrices: one with gene sets found to be significantly UP-regulated and one with gene sets found to be significantly DOWN-regulated

bic-mskcc/bicrnaseq documentation built on May 24, 2019, 3:04 a.m.