A two-step approach to testing overall effect of gene-environment interaction for multiple phenotypes

Area of Science:

• Genetics and Genomics
• Statistical Genetics
• Bioinformatics

Background:

• Gene-environment (GxE) interactions are crucial for many phenotypes but difficult to detect.
• Existing methods often lack sufficient statistical power for robust GxE interaction discovery.
• Simultaneously analyzing multiple phenotypes and employing a two-step approach can enhance power.

Purpose of the Study:

• To develop and evaluate a novel two-step statistical approach for detecting aggregate gene-environment (GxE) effects across multiple phenotypes.
• To improve the power of GxE interaction studies by leveraging multivariate analysis and a sequential testing strategy.

Main Methods:

• Proposed a two-step analysis framework to test for an overall GxE effect across multiple phenotypes.
• Utilized simulations to assess the power gains of the proposed multivariate approach compared to univariate methods.
• Applied the method to UK Biobank data, analyzing three lipid phenotypes (LDL, HDL, Triglyceride) with alcohol consumption as the environmental factor.

Main Results:

• Simulations demonstrated substantial power gains (18-43%) for detecting aggregate GxE effects when multiple phenotypes exhibit GxE pleiotropy.
• The approach successfully identified two loci with an overall GxE effect on the lipid vector in the UK Biobank cohort.
• One significant GxE locus was detected by the proposed method that was missed by competing univariate approaches.

Conclusions:

• The proposed two-step, multivariate approach significantly enhances the power to detect gene-environment interactions across multiple phenotypes.
• This method offers a powerful tool for dissecting complex genetic architectures influenced by environmental factors.
• The R package MPGE is available to facilitate the implementation of this approach in genetic research.