Gene Sert Enrichment Analysis

Usage

gsea(
  rl = NULL,
  gsl = NULL,
  k = 99,
  min_size = 5,
  max_size = 500,
  min_tags = 3,
  abs_val = NULL,
  decreasing = NULL,
  BPPARAMGsl = NULL,
  BPPARAMK = NULL,
  description = NULL,
  out_file_prefix = "gsea_res",
  min.k = 50
)

Arguments

rl: numeric matrix of genes-by-ranking criteria; each column contains numeric values; rownames are mandatory
gsl: named list of gene sets
k: integer, number of permutations
min_size: minimum gene set size
max_size: maximum gene set size
min_tags: minimum number of tags to consider the ES; gene sets with tags < min_tags will be excluded
abs_val: TRUE/FALSE vector that specifies whether to consider absolute values or not in each column of rl; must be of length equal to ncol(rl). If NULL, values will be considered as they are provided
decreasing: TRUE/FALSE vector that specifies whether to order each column of rl decreasingly or not; must be of length equal to ncol(rl). If NULL, all columns will be ranked in decreasing order
BPPARAMGsl: number of cores to use for parallel calculation of gene set lists; the total number of cpu used will be mc_cores_path x mc_cores_perm
BPPARAMK: number of cores to use for parallel calculation of ranked list permutations; the total number of cpu used will be mc_cores_path x mc_cores_perm
description: optional named vector with gene set description; names must be gene seet identifiers
out_file_prefix: prefix for .xlsx and .txt output files
min.k: minimum number of permutations to obtain valid permutation-based statistics

Value

data.frame with: es, enrichment score; nes normalized enrichment score; nperm, number of permutations actually used; p-value, empirical p-value; adjusted p-value, BH FDR; q_val: q-value estimnated from p-values using qvalue package; FDR q-value, empirical FDR; tags, leading edge size; tags_perc, leading edge size percent over gene set; list_top, rank of the ES; list_top_perc, rank of the ES percent over full ranked list; lead_edge, signal strength; lead_edge_subset, gene names of the leading edge, # negative tags, # positive tags, positive tags ratio