|
Abstract:
|
Single nucleotide polymorphism (SNP) set analysis aggregates both common and rare variants and tests for association between a phenotype of interest and a set. However, multiple genes, pathways, or sliding windows are usually investigated across the whole genome, in which all groups are tested separately followed by multiple testing adjustment. We propose a novel method using the devised minorization-maximization (MM) algorithm to prioritize SNP sets in a joint multivariate variance component model and a variance component model with interaction terms. Each SNP set corresponds to a variance component and model selection is achieved by incorporating penalties, including lasso, adaptive lasso, and minimax concave penalty. Simulation studies demonstrate the superiority of our method in model selection performance, as measured by the area under the Precision-Recall curve, compared to the traditional marginal testing methods. Finally, we apply our method to a Merck pharmacogenomics study investigating the genes associated with the Simvastatin and Ezetimibe/Simvastatin responses in the IMPROVE-IT trial. The proposed method is implemented using a Julia package and available for free use.
|
