Detecting Gene-Environment Interaction by Linear Mixed Effects Models

• Activity Number: 350
• Type: Topic Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: International Chinese Statistical Association
• Author(s): Chao Xing* and Hung-Chih Ku and Guan Xing
• Companies: The University of Texas Southwestern Medical Center and The University of Texas Southwestern Medical Center and Gilead Sciences
• Keywords: Gene-environment interaction ; linear mixed models ; generalized F-test

Abstract

Studies of gene-environment interaction are considered pivotal to dissect the multifactorial diseases. Regression-based methods are commonly used to detect gene-environment interaction by assuming a linear relationship between a quantitative trait and both genetic and environmental factors, along with their interactions. However, the assumption of linearity could be violated due to nonlinear responses of genetic variants to environmental stimuli and hence lower the power of detecting gene-environment interaction. In this study we model gene-environment interaction by penalized splines and reparameterize it through a linear mixed effects models with multiple variance components. A generalized F-test of variance components is utilized to test the gene-environment interaction. Further, the algorithm of factored spectrally transformed linear mixed models is employed to increase the computation efficiency. Evaluation of the method and comparison to the existing methods are carried out on both simulated data and a real data example.