limmaToFGSEA: Run gene set enrichment analysis using DE results.

In calebclass/NanoTube: An Easy Pipeline for NanoString nCounter Data Analysis

Run gene set enrichment analysis using DE results.

Description

Use the fgsea library to run gene set enrichment analysis from the Limma analysis results. Genes will be ranked by their log2 fold changes or t-statistics (specified using 'rank.by').

Usage

limmaToFGSEA(
  limmaResults,
  gene.sets,
  sourceDB = NULL,
  min.set = 1,
  rank.by = c("coefficients", "t"),
  skip.first = TRUE
)

Arguments

limmaResults	Result from runLimmaAnalysis.
gene.sets	Gene set file name, in .rds (list), .gmt, or .tab format; or a list object containing the gene sets. Gene names must be in the same form as in the ranked.list.
sourceDB	Source database to include (only if using a .tab-format geneset.file from CPDB).
min.set	Number of genes required to conduct analysis on a given gene set (default = 1). If fewer than this number of genes from limmaResults are included in a gene set, that gene set will be skipped for this analysis.
rank.by	Rank genes by log2 fold changes ('coefficients', default) or t-statistics ('t').
skip.first	Logical: Skip the first factor for gene set analysis? Frequently the first factor is the 'Intercept', which is generally uninteresting for GSEA (default TRUE).

Details

Limma returns matrices of coefficients and t statistics with columns for each column in the design matrix. This function will conduct a separate enrichment analysis on each column from the relevant matrix. Because the first column may be an "intercept" term, which is generally not relevant for enrichment analysis, the user may want to skip analysis for that term (using skip.first = TRUE, the default).

Value

A list containing data frames with the fgsea results for each comparison.

Examples

data("ExamplePathways")
data("ExampleResults") # Results from runLimmaAnalysis

# Use the default settings
fgseaResults <- limmaToFGSEA(ExampleResults, gene.sets = ExamplePathways)

# Only include gene sets with at least 5 genes in the NanoString data set,
# and rank genes by their "t" statistics.
fgseaResults <- limmaToFGSEA(ExampleResults, gene.sets = ExamplePathways,
                             min.set = 5, rank.by = "t")

calebclass/NanoTube documentation built on Nov. 21, 2023, 12:31 p.m.