checkModules: Evaluate a set of modules using gene annotations

Description Usage Arguments Details Value Examples

Description

Evaluates a set of gene modules (e.g., the output from predictModules) by checking them for enrichment of gene annotations.

Usage

1
checkModules(x, a, p = 0.05, return.p.matrix = FALSE)

Arguments

x

a binary matrix indicating module membership (genes in rows, modules in columns)

a

a binary gene annotation matrix (genes in rows, annotations in columns)

p

FDR-adjusted p-value below which to consider an annotation significant

return.p.matrix

logical value indicating whether the full p-value matrix should be returned in the output

Details

This function evaluates a set of predicted gene modules by testing whether each module is enriched for any gene annotations. Enrichment p-values are calculated using hypergeometric statistics (see phyper). If x is not a binary matrix, module membership is determined using partition with default paramaters. The FDR-correction (see p.adjust) is applied to all p-values prior to determining signficance.

Value

A list with the following elements:

num.annotations

Number of annotations tested (number of columns in a)

num.modules

Number of modules tested (number of columns in x)

anns.signif

Number of annotations that were significant in at least one module

mods.signif

Number of modules that were significant for at least one annotation

p.matrix

The module-by-annotation p-value matrix (if return.p.matrix == TRUE)

Examples

1
2
3
4
5
6
7
x = matrix(rnorm(100), 10, 10)
a = matrix(sample(c(0,1), size = 100, replace = TRUE), 10, 10)
m = predictModules(x)$S
rownames(a) = rownames(m) = as.character(1:10)
m.p = partition(m, t = 2)
m.p2 = splitMatrix(m.p)
checkModules(m.p2, a)

MPCary/DEXICA documentation built on May 4, 2019, 2:35 p.m.