Penalized Regression and Risk Prediction in Genome-Wide Association Studies

• Activity Number: 559
• Type: Topic Contributed
• Date/Time: Wednesday, August 1, 2012 : 2:00 PM to 3:50 PM
• Sponsor: IMS
• Author(s): Erin Edward Austin*+ and Wei Pan and Xiaotong Shen
• Companies: University of Minnesota and University of Minnesota and University of Minnesota-Twin Cities
• Keywords: Penalized Regression ; Risk Prediction

Abstract

An important task in personalized medicine is risk prediction based on a large number of single-nucleotide polymorphisms (SNPs) with data collected from genome-wide association studies (GWAS). Penalized regression equipped with variable selection, such as Lasso, is deemed to be promising in this setting. However, the sparsity assumption taken by the Lasso and many other penalized regression techniques may not be applicable here: it is now hypothesized that many common diseases are associated with many common SNPs with small to moderate effects. In this project, we use the GWAS data from the Wellcome Trust Case Control Consortium (WTCCC) to investigate the performance of various unpenalized and penalized regression approaches with sparse or non-sparse models.