Bayesian Hierarchical Structured Variable Selection Methods with Application to Mip Studies in Breast Cancer

• Activity Number: 384
• Type: Topic Contributed
• Date/Time: Tuesday, August 5, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Physical and Engineering Sciences
• Author(s): Lin Zhang*+ and Veera Baladandayuthapani and Bani K. Mallick and Ganiraju C. Manyam and Patricia A. Thompson and Melissa L. Bondy and Kim-Ahn Do
• Companies: MD Anderson Cancer Center and MD Anderson Cancer Center and Texas A&M and MD Anderson Cancer Center and University of Arizona and Baylor University and MD Anderson Cancer Center
• Keywords: copy number alteration ; hierarchical variable selection ; lasso ; MIP data ; MCMC

Abstract

The analysis of copy number alterations has been a focus of research to identify genetic markers of cancer. One high-throughput technique recently adopted is the use of molecular inversion probes (MIPs), whose resulting data consist of high-dimensional copy number profiles that can be used to ascertain probe-specific copy number alterations in correlative studies with patient outcomes to guide risk stratification and future treatment. We propose a novel Bayesian variable selection method, the hierarchical structured variable selection (HSVS) method, which accounts for the natural gene and probe-within-gene architecture to identify important genes and probes associated with clinically relevant outcomes. The HSVS model utilizes a discrete mixture prior distribution for group selection and group-specific Bayesian lasso hierarchies for variable selection within groups. We provide methods for accounting for serial correlations within groups that incorporate Bayesian fused lasso methods for within-group selection. Through simulations we establish that our method results in lower model errors than other methods when a natural grouping structure exists. We apply our method to an MIP study