A0311
Title: Advances in identifying latent gene-gene and gene-environment interactions for binary outcomes
Authors: Lei Sun - University of Toronto (Canada) [presenting]
Abstract: Investigating gene-environment (GxE) interactions without direct environmental data poses challenges, and exhaustive gene-gene (GxG) searches face issues with large-scale multiple-hypothesis testing. These complexities are explored, focusing on binary traits. For continuous traits, latent interactions induce artificial heteroskedasticity, which is detectable with methods like Levene's test. However, binary traits present unique challenges due to their variance being determined by the mean. These challenges are addressed, and a novel approach is proposed and evaluated based on non-additive genetic effects. Practical insights are demonstrated through an application to the UK Biobank data.
