Genetic association analysis of coronary heart disease by profiling gene-environment interaction based on latent components in longitudinal endophenotypes.

Studies of complex diseases collect panels of disease-related traits, also known as secondary phenotypes or endophenotypes. They reflect intermediate responses to environment exposures, and as such, are likely to contain hidden information of gene-environment (G x E) interactions. The information can be extracted and used in genetic association studies via latent-components analysis. We present such a method that extracts G x E information in longitudinal data of endophenotypes, and apply the method to repeated measures of multiple phenotypes related to coronary heart disease in Genetic Analysis Workshop 16 Problem 2. The new method identified many genes, including SCNN1B (sodium channel nonvoltage-gated 1 beta) and PKP2 (plakophilin 2), with potential time-dependent G x E interactions; and several others including a novel cardiac-specific kinase gene (TNNI3K), with potential G x E interactions independent of time and marginal effects.