eQuIPS: eQTL Analysis Using Informed Partitioning of SNPs - A Fully Bayesian Approach.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Expression quantitative trait locus (eQTL) studies aim to link genetic variations to gene expression.
• Identifying significant single nucleotide polymorphisms (SNPs) is crucial for understanding gene regulation.
• Current methods may not fully leverage functional information for SNP prioritization.

Purpose of the Study:

• To develop a Bayesian approach integrating functional significance scores into SNP prior effect sizes for eQTL studies.
• To improve the objective identification and ranking of causal SNPs.
• To enhance the efficiency of follow-up studies in genetic research.

Main Methods:

• Development of a Bayesian multi-SNP Markov chain Monte Carlo (MCMC) approach.
• Introduction of a Normal Gamma prior to incorporate functional information.
• Partitioning SNPs into functional groups with group-specific prior distributions based on significance scores.
• Testing on simulated datasets and human eQTL data with validated causal SNPs.

Main Results:

• The modified Normal Gamma prior method consistently performed as well as or better than other methods in simulations.
• In human eQTL data analysis, the new method dramatically improved the ranks of validated causal SNPs.
• Three out of four validated SNPs were ranked in the top 1% and one in the top 2% using the new method, compared to top 1%, 4%, 20%, and 59% for the standard Normal Gamma.

Conclusions:

• The developed Bayesian method effectively integrates functional information to enhance SNP prioritization in eQTL studies.
• This approach significantly increases the likelihood of identifying true causal SNPs for experimental follow-up.
• The method offers a substantial improvement over existing approaches for eQTL analysis and genetic discovery.