Adaptive P-Value Thresholding for Multiple Hypothesis Testing
with Side Information

Implementation of adaptive p-value thresholding (AdaPT), including both a framework that allows the user to specify any
algorithm to learn local false discovery rate and a pool of convenient functions that implement specific
algorithms. See Lei, Lihua and Fithian, William (2016) .

Overview

This package implements Adaptive P-Value Thresholding in the paper: AdaPT: An interactive procedure for multiple testing with side information. It includes both a framework that allows the user to specify any algorithm to learn local FDR and a pool of convenient functions that implement specific algorithms:

adapt() provides a generic framework of AdaPT permitting any learning algorithm;

adapt_glm(), adapt_gam() and adapt_glmnet() provide convenient wrappers of AdaPT using Generalized Linear Models (GLM), Generalized Additive Models (GAM) and L1-penalized GLMs;

Install the adaptMT package then read vignette("adapt_demo", package = "adaptMT").

If one wants to access the vignette, run the following code to build the vignette. This might update other related packages and please be patient if so.