A fast and powerful linear mixed model approach for genotype-environment interaction tests in large-scale GWAS

Area of Science:

• Human genetics
• Statistical genetics
• Genomics

Background:

• Genotype-by-environment interaction (GEI) is vital for understanding complex human traits.
• Detecting GEI signals in large-scale genome-wide association studies (GWAS) is challenging due to population stratification and sample relatedness.

Purpose of the Study:

• To develop a fast and powerful method for detecting GEI effects in large-scale GWAS.
• To introduce fastGWA-GE, a linear mixed model-based approach for GEI and G+GEI joint effect testing.

Main Methods:

• Proposed fastGWA-GE, a linear mixed model-based approach.
• Conducted extensive simulations to evaluate performance against existing GEI test methods.
• Applied fastGWA-GE to UK Biobank GWAS data (~7.27 million variants, 452,249 individuals) for 13 quantitative traits and 5 environment variables.

Main Results:

• fastGWA-GE demonstrated superior performance in simulations, showing better genomic inflation control, increased power, and significantly faster runtimes.
• Identified 96 significant GEI signals (P < 1x10^-9) in the UK Biobank data.
• Discovered 27 novel GEI associations, underscoring the method's effectiveness.

Conclusions:

• fastGWA-GE is an effective and efficient tool for discovering GEI signals in large-scale GWAS.
• The method facilitates a deeper understanding of how genetic and environmental factors interact to influence complex human traits.