Run gene set enrichment analysis on MetaVolcano results

This function performs fast gene set enrichment analysis (fGSEA) on the results of a random-effects model meta-analysis. Three ranking strategies are supported: by summary fold-change, by signed significance, or by a weighted combination of both. Gene sets can be provided directly or downloaded automatically from MSigDB via msigdbr.

enrichment_mv(
  remres,
  pathways = NULL,
  species = "Homo sapiens",
  category = "C5",
  subcategory = NULL,
  ranking = "fc",
  minSize = 15,
  maxSize = 500,
  nPermSimple = 10000,
  plot_top_n = 30,
  plot_padj = 0.1,
  clean_pathway_names = TRUE,
  colors = c(down = "#083e46", up = "#811820"),
  plot_title = NULL,
  seed = 42
)

Arguments

remres: MetaVolcano object. Output of the rem_mv() function <MetaVolcano>
pathways: named list of gene sets (character vectors). If NULL, gene sets are automatically downloaded via msigdbr using species, category, and subcategory parameters
species: species name for msigdbr. Default "Homo sapiens"
category: MSigDB category: "H" (hallmarks), "C2" (curated), "C5" (GO), "C7" (immunologic), etc. Default "C5"
subcategory: optional MSigDB subcategory: "GO:BP", "GO:CC", "GO:MF", "CP:KEGG", "CP:REACTOME", etc. Default NULL (all)
ranking: method for ranking genes: "fc" uses the REM summary fold-change; "signed_p" uses -log10(p) * sign(FC); "weighted_fc" uses FC * -log10(p) <string>
minSize: minimum gene set size to test <integer>
maxSize: maximum gene set size to test <integer>
nPermSimple: number of permutations for p-value estimation <integer>
plot_top_n: number of top pathways to plot <integer>
plot_padj: adjusted p-value threshold for plotting. Default 0.1
clean_pathway_names: logical. If TRUE, removes database prefixes (e.g. GOBP_, KEGG_) and replaces underscores with spaces. Default TRUE
colors: vector of 2 colors for the plot c(down, up)
plot_title: custom plot title. If NULL, auto-generated
seed: random seed for reproducibility <integer>

Value

list with two elements: result (data.frame with fGSEA results) and plot (ggplot2 lollipop plot)

Examples

if (FALSE) { # \dontrun{
data(diffexplist)
mv <- rem_mv(diffexplist, metathr = 0.01)

# Using automatic MSigDB download (GO Biological Process)
enrich <- enrichment_mv(mv, subcategory = "GO:BP")
enrich$plot

# Using Hallmarks
enrich <- enrichment_mv(mv, category = "H")

# Using KEGG
enrich <- enrichment_mv(mv, category = "C2", subcategory = "CP:KEGG")

# Using custom pathways
library(msigdbr)
my_paths <- split(
    msigdbr(species = "Homo sapiens", category = "C5")$gene_symbol,
    msigdbr(species = "Homo sapiens", category = "C5")$gs_name
)
enrich <- enrichment_mv(mv, pathways = my_paths)
head(enrich$result)
} # }