Mixture model-based association analysis with case-control data in genome wide association studies

Area of Science:

• Genetics
• Statistical genetics
• Computational biology

Background:

• Multilocus haplotype analysis combined with genome-wide association studies (GWAS) can detect disease associations missed by individual loci.
• Identifying disease-associated haplotypes becomes difficult with a large number of candidate variants.
• Existing methods often lack direct linkage to haplotype disease penetrance, limiting their efficiency.

Purpose of the Study:

• To propose an efficient method for detecting disease risk haplotypes.
• To address the limitations of current approaches by directly incorporating disease penetrance.

Main Methods:

• Development of a mixture model-based approach for haplotype clustering based on estimated disease penetrances.
• Introduction of a hypothesis test for underlying haplotype inheritance patterns.
• Evaluation of the approach using simulations and real genetic data.

Main Results:

• The proposed mixture model effectively clusters haplotypes according to their disease penetrance.
• The new method demonstrates superior performance compared to an existing multiple testing approach.
• Theoretical justification and validation through simulations and real data analysis support the model's efficacy.

Conclusions:

• The proposed mixture model offers a more efficient and penetrance-linked strategy for identifying risk haplotypes.
• This approach enhances the analysis of complex genetic associations in disease studies.
• The method provides a valuable tool for genetic epidemiology and personalized medicine.