Authors
Kun Fan, Jie Ren, Shuangge Ma, Cen Wu
Published in
Statistics in medicine. Volume 45. Issue 25-27.
Abstract
In high-dimensional longitudinal interaction studies, it is challenging to perform variable selection and statistical inference when the assumed interaction model is misspecified under heterogeneous disease phenotypes. Partially motivated by analyzing a subset of the Childhood Asthma Management Program (CAMP) data with high-dimensional single nucleotide polymorphism (SNP) measurements, we develop a novel sparse robust Bayesian mixed model to capture nonlinear longitudinal interactions. In terms of both parameter estimation and statistical inference, the proposed mixed model is robust not only to outliers and heavy-tailed distributions in the response variable but also to misspecification of nonlinear interaction effects. We have developed a Gibbs sampler incorporating spike-and-slab priors to promote exact sparsity in the identification of appropriate forms of main and interaction effects. Simulation results reveal superior performance of the proposed fully robust Bayesian analysis in identification, estimation, and statistical inference compared with alternative methods. In particular, the proposed analysis enables exact statistical inference in misspecified interaction models by yielding Bayesian credible intervals with nominal coverage probabilities on finite samples. Application of the proposed and alternative methods to the CAMP study sheds novel insight into asthma etiology.
PMID:
43046303
Bibliographic data and abstract were imported from PubMed on 18 Nov 2026.
Read full publication at:
Please sign in
to see all details.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 7
- Comments 0