MixedBayes: An R Package for Longitudinal Gene–Environment Interaction Analysis Using Robust Sparse Bayesian Mixed Models

Robust variable selection methods have emerged as powerful tools for dissecting high-dimensional gene–environment interactions in longitudinal studies, owing to their ability to accommodate intra-cluster correlations, capture structured sparsity, and handle heavy-tailed repeated measures. Despite these advantages, variable selection-based interaction analysis still suffers from a lack of valid inferential tools to quantify the uncertainty associated with important gene–environment interactions. In this paper, we introduce the R package mixedBayes (version 0.2.5), which implements fully Bayesian robust mixed-effects models proposed in recent work for high-dimensional longitudinal gene–environment interaction analysis. Specifically, the package considers two major classes of mixed models. The first accommodates interactions between omics features and treatment effects arising from repeated-measures one-way ANOVA with high-dimensional genetic factors. The second provides a more general framework for modeling interactions between individual genetic main effects and environmental factors. Both models enable posterior Bayesian inference via Markov chain Monte Carlo (MCMC). We provide detailed numerical examples and accompanying R code to facilitate robust interaction analysis using mixedBayes. In addition, a case study based on longitudinal asthma data with high-dimensional SNP measurements is presented.