lfdr.bbe: Binomial-based LFDR estimator.

Description

A binomial-based estimator of the local false discovery rate (BBE) with p0 estimation (use lfdr.bbe). When p0 = 1 the estimator is called BBE1 (use lfdr.bbe1).

Usage

lfdr.bbe(pvalue, p0 = NULL, robust = FALSE, monotonic = FALSE, ...) 
lfdr.bbe1(pvalue, robust = FALSE, monotonic = FALSE, ...)

Arguments

pvalue

Input numeric vector of p-values.

Proportion of non-affected features, it is NULL or a value within [0,1]. For lfdr.bbe1 p0=1, while for lfdr.bbe it can be set or it can be estimated internally by p0est, setting p0=NULL (the default value).

robust

Logical. If robust=TRUE, the estimate is more robust for small p-values. If robust=TRUE, the estimate is more robust for small p-values (see package qvalue).

monotonic

Logical. If monotonic=TRUE, LFDR estimates are monotonically increasing with p-values.

...

Further arguments passed to function p0est.

Value

LFDR.hat: Vector of estimates of the LFDR,
p0.hat: Estimated p0 (for BBE), 1 (for BBE1)
pvalue: Vector of p-values.
info: Method name and information about computation failure.

References

Bickel, D. R. (2013). Simple estimators of false discovery rates given as few as one or two p-values without strong parametric assumptions. Statistical Applications in Genetics and Molecular Biology, Statistical Applications in Genetics and Molecular Biology, 12(4), pp. 529-43.

Padilla, M., & Bickel, D. R. (2012). Estimators of the local false discovery rate designed for small numbers of tests. Statistical Applications in Genetics and Molecular Biology, 11(5), art. 4.

Examples

Run this code

# a p-values vector with missing elements
pp <- runif(15) ;pp[c(13,2,5)]<-NA

#BBE1:
z1 <- lfdr.bbe1(pvalue = pp, robust = TRUE)
z2 <- lfdr.bbe1(pvalue = pp, robust = FALSE, monotonic = TRUE)

#estimating or setting p0:
z3 <- lfdr.bbe(pvalue = pp, p0 = NULL)
z4 <- lfdr.bbe(pvalue = pp, pi0.method = "bootstrap")
z5 <- lfdr.bbe(pvalue = pp, p0 = 0.8, robust = TRUE, monotonic = TRUE)

Run the code above in your browser using DataLab

Last chance! 50% off unlimited learning