FDR Correction (Benjamini-Hochberg)

Wrapper around stats::p.adjust for FDR correction using the Benjamini-Hochberg procedure. Provides interface similar to MATLAB fdr_bh.

Usage

fdr_bh(pvals, q = 0.05, method = "BH")

Arguments

pvals

numeric vector of p-values.

q

desired false discovery rate (default 0.05).

method

adjustment method passed to p.adjust. Default "BH".

Value

list with:

h

logical vector indicating significant tests after FDR correction

crit_p

critical p-value threshold (largest p-value deemed significant)

adj_p

adjusted p-values

References

Benjamini Y, Hochberg Y (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society B, 57, 289-300.

Examples

pvals <- c(0.001, 0.01, 0.03, 0.04, 0.05, 0.1, 0.5)
result <- fdr_bh(pvals, q = 0.05)
result$h
#> [1]  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE