A unified approach to false discovery rate estimation

BMC Bioinformatics

Table 2 Overview over some commonly used FDR estimation procedures.

Name	FDR Type	Input Data	Comments	Description of Algorithm
fdrtool	fdr, Fdr	p-values,	Modified Grenander density,	This paper
		z-scores,	estimate, empirical null model,
		t-scores, correlations.	selection of truncation point by FNDR.
BUM	fdr, Fdr	p-values	Completely parametric model.	[27]
SAGx	fdr, Fdr	p-values	Grenander density estimate.	[5, 6]
qvalue	Fdr	p-values	Diverse estimates of η₀ available.	[12, 7]
nFDR	Fdr	p-values	Bernstein polynomial density.	[24]
multtest	Fdr	p-values	Benjamini-Hochberg algorithm.	[2]
LBE	Fdr	p-values	Location-based estimator.	[32]
locfdr	fdr	z-scores	Regression spline density estimate, empirical null model.	[4, 28]
nomi	fdr	z-values	Normal mixture modeling of non-null density.	[22]
LocalFDR	fdr	p-values	Uses loess smoothing.	[15]
kerfdr	fdr	p-values	Kernel density estimator.	[23]
twilight	fdr	p-values	KS fit of truncation point.	[30]
localFDR	fdr	p-values	Based on stochastic order model.	[33]

Note: The first column refers to an R package or R script implementing the respective FDR algorithm.

ISSN: 1471-2105