Efficient and accurate framework for genome-wide gene-environment interaction analysis in large-scale biobanks

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Gene-environment interaction (G×E) is crucial for understanding complex traits.
• Genome-wide association studies (GWAS) are expanding to complex phenotypes.
• Existing scalable methods often lack support for diverse trait types.

Purpose of the Study:

• To develop a scalable and accurate framework (SPAGxE_CCT) for G×E analysis across diverse trait types.
• To extend the framework (SPAGxEmix_CCT) to account for population stratification in multi-ancestry populations.
• To enable identification of ancestry-specific G×E effects using local ancestry (SPAGxEmix_CCT-local).

Main Methods:

• Developed SPAGxE_CCT, a genotype-independent model using saddlepoint approximation (SPA) for precise p-value calculation.
• Extended SPAGxE_CCT to SPAGxEmix_CCT for population stratification and multi-ancestry data.
• Further extended to SPAGxEmix_CCT-local for local ancestry-based G×E analysis.

Main Results:

• SPAGxE_CCT and SPAGxEmix_CCT demonstrate scalability for large-scale cohort analysis.
• The methods effectively control type I error rates.
• The framework maintains statistical power in G×E analyses.

Conclusions:

• SPAGxE_CCT and its extensions provide a robust and scalable solution for G×E analysis in diverse populations and trait types.
• These methods enhance the ability to dissect complex genetic architectures involving environmental factors.
• The framework is validated through simulations and real-world UK Biobank data analysis.