A Bayesian hierarchically structured prior for rare-variant association testing.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) are crucial for identifying complex disease-genetic variant associations.
• Standard single-variant analyses lack power for rare genetic variants.
• Set-based methods enhance power by aggregating information from multiple rare variants.

Purpose of the Study:

• To propose and evaluate adaptive hierarchically structured variable selection (HSVS-A) for rare variant association testing.
• To estimate individual rare variant effects and boost statistical power.
• To apply HSVS-A to identify pathways associated with Crohn's disease.

Main Methods:

• Development of the HSVS-A method for set-based rare variant analysis.
• Integration of a pairwise weighting scheme within HSVS-A to correlate variants by significance.
• Simulation studies for continuous and dichotomous phenotypes.
• Application of HSVS-A to Crohn's disease data from the Wellcome Trust Case Control Consortium.

Main Results:

• HSVS-A demonstrates high power with multiple causal rare variants, even with non-causal variants present.
• The method performs effectively for both continuous and dichotomous phenotypes.
• Two pathways were identified harboring novel protective rare variants for Crohn's disease.

Conclusions:

• HSVS-A is a powerful tool for rare variant association studies, outperforming standard methods.
• The method's adaptive weighting scheme enhances power by leveraging information from correlated variants.
• HSVS-A successfully identified novel genetic pathways associated with Crohn's disease.