High-Dimensional Gene-Environment Interaction Analysis

Area of Science:

• Genetics
• Environmental Health
• Biostatistics

Background:

• Complex diseases arise from genetic and environmental factors, with gene-environment (G-E) interactions playing a significant role.
• Current G-E interaction analyses often employ supervised frameworks for genetic and environmental factors in relation to disease.
• A statistical perspective is needed to review methodological advancements in G-E interaction analysis.

Purpose of the Study:

• To provide a selective review of statistical methodologies for gene-environment interaction analysis.
• To categorize and discuss the main frameworks and techniques used in G-E interaction studies.
• To highlight considerations for applying these methods in diverse research scenarios.

Main Methods:

• Review of hypothesis testing, variable selection, and dimension reduction techniques.
• Discussion of testing-based, estimation-based, and prediction-based analytical frameworks.
• Exploration of linear/nonlinear, fixed/random effects, marginal/joint, and Bayesian/frequentist analyses.

Main Results:

• Identified three primary statistical frameworks: testing-based, estimation-based, and prediction-based.
• Detailed various analytical approaches including linear/nonlinear and Bayesian/frequentist methods.
• Highlighted statistical properties, computational aspects, and practical applications of G-E interaction methods.

Conclusions:

• Methodological diversity exists for analyzing gene-environment interactions, catering to different research goals.
• The review facilitates the appropriate application of statistical techniques for G-E interaction analysis.
• Future research directions in statistical G-E interaction analysis are outlined.