Penalized Regression and Prediction of Disease Outcomes

• Activity Number: 589
• Type: Topic Contributed
• Date/Time: Wednesday, August 7, 2013 : 2:00 PM to 3:50 PM
• Sponsor: ENAR
• Author(s): Erin Austin*+ and Wei Pan and Xiaotong Shen
• Companies: University of Minnesota - Division of Biostatistics and University of Minnesota and University of Minnesota
• Keywords: Penalized Regression ; TLP ; LASSO ; Elastic Net ; GWAS ; Rare Variant

Abstract

An important task in genetics-based personalized medicine is prediction of disease outcomes. Penalized regression is particularly promising in this area because it can leverage information about a disease's genetic architecture to improve estimation. In this project, we initially 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. Next, we utilize the sequence data from the Genetic Analysis Workshop 18 (GAW18) to demonstrate how penalized regression can use rare variants to improve estimation while accounting for environmental covariates.