Testing and estimating gene-environment interactions in family-based association studies

Area of Science:

• Genetics
• Biostatistics
• Epidemiology

Background:

• Analyzing gene-environment interactions is crucial for understanding complex diseases.
• Family-based studies offer unique advantages but face challenges like population stratification.
• Existing methods may be susceptible to confounding from unmeasured factors.

Purpose of the Study:

• To develop robust and efficient statistical tests and estimators for gene-environment/gene-drug interactions in family-based association studies.
• To ensure these methods are resilient to confounding from population admixture and stratification.
• To apply and validate the methodology in a real-world disease context.

Main Methods:

• Utilized causal inference methodology for developing statistical tests.
• Incorporated haplotype analysis, dichotomous/quantitative phenotypes, and complex exposure variables.
• Employed simulation studies and a case study of chronic obstructive pulmonary disease (COPD).

Main Results:

• The proposed tests and estimators demonstrated robustness against unmeasured confounding under specific conditions (Mendel's law, exposure independence from gene).
• Data analysis identified a potential gene-environment interaction between a Serpine2 single nucleotide polymorphism and smoking in a COPD study.
• Simulation studies confirmed sufficient statistical power and validity of the methods for realistic sample sizes.

Conclusions:

• The developed methodology provides a robust framework for gene-environment interaction analysis in family studies, even with population structure.
• The findings highlight the importance of considering gene-environment interactions, such as with Serpine2 and smoking, in disease etiology.
• This approach enhances the reliability of genetic association studies by addressing confounding factors.