Skip to content

Latest commit

 

History

History
43 lines (28 loc) · 788 Bytes

README.md

File metadata and controls

43 lines (28 loc) · 788 Bytes

swfdr

Science-wide false discovery rate

swfdr is an R package for dealing with multiple hypothesis testing.

Installation

The package is available through Bioconductor.

library(BiocManager)
BiocManager::install("swfdr")

It is also possible to install in-development versions from github.

library(devtools)
install_github("leekgroup/swfdr")

Example

A minimal working example for using the lm_pi0 and lm_qvalue functions is as follows

library(swfdr)
x <- c(rbeta(1000, 0.2, 1), runif(1000, 0, 1))
x.covariate = rep(c(1,2), each=1000)
# compute covariate-adjusted pi0 - probability of null hypothesis to be true
x.pi0 = lm_pi0(x, X=x.covariate)
x.pi0
# compute covariate-adjusted qvalues
x.qvalues <- lm_qvalue(x, X=x.covariate)
x.qvalues