A model for fine mapping in family based association studies

Area of Science:

• Genetics
• Biostatistics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) identify genetic regions linked to complex diseases.
• Further characterization of these regions is crucial for understanding disease mechanisms.

Purpose of the Study:

• To propose a latent class model for evaluating candidate genetic regions.
• To estimate linkage disequilibrium (LD) between observed markers and unobserved disease loci.
• To assess the impact of disease alleles on disease risk within families.

Main Methods:

• Development of a latent class model incorporating observed genetic markers and an unobserved disease locus.
• Estimation of the joint distribution of alleles and a penetrance parameter.
• Inclusion of family-specific random effects to account for varying disease prevalences.
• Utilizing a likelihood framework for model estimation and simulation-based property assessment.

Main Results:

• The proposed model successfully estimates LD and penetrance parameters.
• Simulations demonstrate the model's robust properties.
• Application to an Alzheimer's dataset confirmed known findings in the ApoE region.

Conclusions:

• The latent class model provides a powerful framework for fine-mapping genetic regions associated with complex diseases.
• This approach enhances the understanding of genetic architecture and disease risk.
• The model is applicable to real-world genetic datasets, such as for Alzheimer's disease.