Robust Bayesian Variable Selection for Gene-environment Interactions
Gene-environment (G×E) interactions have important implications to elucidate the etiology of complex diseases beyond the main genetic and environmental effects. Outliers and data contamination in disease phenotypes of G×E studies have been commonly encountered, leading to the development of a broad spectrum of robust penalization methods. Nevertheless, within the Bayesian framework, the issue has not been taken care of in existing studies. We develop a robust Bayesian variable selection method for G×E interaction studies. The proposed Bayesian method can effectively accommodate heavy–tailed errors and outliers in the response variable while conducting variable selection by accounting for structural sparsity. In particular, the spike–and–slab priors have been imposed on both individual and group levels to identify important main and interaction effects. An efficient Gibbs sampler has been developed to facilitate fast computation. The Markov chain Monte Carlo algorithms of the proposed and alternative methods are efficiently implemented in C++.
- To install from github, run these two lines of code in R
install.packages("devtools")
devtools::install_github("jrhub/roben")
- Released versions of roben are available on CRAN (link), and can be installed within R via
install.packages("roben")
library(roben)
data(GxE_small)
iter = 5000
fit=roben(X, Y, E, clin, iterations = iter)
fit$coefficient
## Ture values of parameters of mian G effects and interactions
coeff$GE
## Compute TP and FP
sel = GxESelection(fit)
pos = which(sel$indicator != 0)
tp = length(intersect(which(coeff$GE != 0), pos))
fp = length(pos) - tp
list(tp=tp, fp=fp)
fit=roben(X, Y, E, clin, iterations = iter, robust=FALSE)
sel = GxESelection(fit)
pos = which(sel$indicator != 0)
tp = length(intersect(which(coeff$GE != 0), pos))
fp = length(pos) - tp
list(tp=tp, fp=fp)
This package provides implementation for methods proposed in
- Ren J, Zhou F, Li X, Ma S, Jiang Y, Wu C. Robust Bayesian variable selection for gene-environment interactions. Biometrics. 2023 Jun;79(2):684-694. doi: 10.1111/biom.13670. Epub 2022 Apr 16. PMID: 35394058.