Inference on haplotype/disease association using parent-affected-child data: the projection conditional on parental...

Area of Science:

• Genetics
• Epidemiology
• Statistical genetics

Background:

• Log-linear models are used to analyze disease risk based on genetic data.
• Case-parent trio data is valuable for genetic association studies.
• Population stratification and environmental factors can confound genetic analyses.

Purpose of the Study:

• To develop a robust statistical method for parameter inference in log-linear models of relative disease risk.
• To analyze case-parent trio data, accounting for population stratification and environmental interactions.
• To generalize and improve upon existing methods for genetic association studies.

Main Methods:

• Development of a novel inference method for log-linear models.
• Application to case-parent trio data.
• Incorporation of robustness to population stratification.
• Inclusion of haplotype-environment interactions and handling of missing genotype data.

Main Results:

• The proposed method provides reliable parameter inference for disease risk.
• The method is robust to population stratification.
• It allows for the analysis of haplotype-environment interactions.
• It generalizes and improves upon previous methods, including handling missing data.

Conclusions:

• The developed method offers a robust and flexible approach for genetic association studies using case-parent trio data.
• It addresses limitations of previous methods, particularly regarding population stratification and environmental factors.
• The approach enhances the accuracy of disease risk parameter estimation in genetic epidemiology.