Skip to contents

GSEA framework

Source: R/framework.R
gsea_framework.Rd

GSEA framework

Usage

gene_level_log2fc(mat, condition)

gene_level_s2n(mat, condition)

gene_level_tvalue(mat, condition)

gene_level_sam(mat, condition)

gene_level_log_p(mat, condition)

set_level_mean(s, l)

set_level_median(s, l)

set_level_maxmean(s, l)

set_level_wilcox(s, l)

set_chi_square(s, l)

set_level_ks(s, l)

Arguments

mat

The input matrix.

condition

The condition labels of samples. The value should be a factor where level[1] corresponds to the treatment group and level[2] corresponds to the control group.

s

A vector of gene scores.

l

A logical vector representing whether genes in s are in a specific gene set.

Details

There are the following gene-level statistics

  • log2fc

  • signal to noise ratio

  • t-value

  • SAM t-value

  • -log(p)

There are the following gene set level statistics:

  • mean

  • median

  • maxmean

  • wilcox ranksum

  • ks. Since we only support one-side test, the ES is defined as max(abs(f2 - f1)).

  • chi-square. Only works when gene-level scores are binary.