A hidden Markov random field model for genome-wide association studies

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) are crucial for identifying genetic variants linked to complex traits.
• Current GWAS analysis methods, like single nucleotide polymorphism (SNP) analysis, often lack statistical power due to conservative error control from Bonferroni correction, especially when SNPs are in linkage disequilibrium (LD).

Purpose of the Study:

• To propose a novel Hidden Markov Random Field (HMRF) model for GWAS analysis.
• To improve statistical power and accuracy in identifying disease-associated genetic variants by effectively utilizing LD information.

Main Methods:

• Developed a Hidden Markov Random Field (HMRF) model incorporating prior linkage disequilibrium (LD) information via a weighted LD graph.
• Employed an efficient iterative conditional mode algorithm for model parameter estimation.
• Utilized posterior probabilities derived from the HMRF model to implement a false discovery rate controlling procedure for SNP selection.

Main Results:

• Simulation studies showed increased statistical power compared to single SNP analysis.
• The HMRF model effectively identified SNPs with borderline significance that were in high LD with significant SNPs.
• The method reduced false positive identifications by simultaneously considering SNPs in LD.
• Applied to neuroblastoma GWAS data, the model identified one novel potentially associated SNP.

Conclusions:

• The proposed HMRF model offers a powerful and accurate approach for GWAS analysis.
• Leveraging LD information within the HMRF framework enhances the detection of disease-associated genetic variants.
• This method provides a valuable tool for genetic research, particularly for complex traits.