Gene-environment interaction testing in family-based association studies with phenotypically ascertained samples: a...

Area of Science:

• Genetics
• Epidemiology
• Biostatistics

Background:

• Gene-environment (G × E) interactions are crucial for understanding complex traits.
• Family-based studies are valuable for dissecting genetic and environmental influences.
• Phenotypic ascertainment criteria can introduce bias in G × E interaction analyses.

Purpose of the Study:

• To develop a novel statistical method for testing G × E interactions in family-based studies with ascertainment.
• To address challenges including unmeasured confounding and incomplete parental genotypes.
• To provide a robust approach for analyzing G × E effects on complex traits.

Main Methods:

• Employed G-estimation, a semiparametric technique from causal inference.
• Developed a test that accounts for ascertainment conditions and population substructure.
• Allowed for incomplete parental genotype data.

Main Results:

• The proposed G-estimation method demonstrated robustness in simulation studies.
• The method was compared against conditional likelihood and QBAT-I tests.
• Applied the approach to a real-world study of chronic obstructive pulmonary disorder.

Conclusions:

• The novel G-estimation approach provides a robust method for testing G × E interactions in ascertainment-based family studies.
• This method effectively handles unmeasured confounding and incomplete genetic data.
• The findings have implications for genetic epidemiology and complex trait research.