Combining eQTL and SNP Annotation Data to Identify Functional Noncoding SNPs in GWAS Trait-Associated Regions

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Genome-Wide Association Studies (GWAS) identify trait-associated genomic regions.
• Prioritizing causal variants within these regions, especially noncoding single nucleotide polymorphisms (SNPs), remains challenging.
• Noncoding SNPs may influence gene regulation and trait association through cis-expression quantitative trait loci (cis-eQTLs) and regulatory potential (rSNP).

Purpose of the Study:

• To develop and present a statistical method for prioritizing candidate causal noncoding SNPs identified through GWAS.
• To integrate multiple data types, including GWAS association p-values, cis-eQTL association p-values, and rSNP prediction scores, into a unified framework.
• To enable secondary analysis of existing GWAS summary data for improved variant prioritization.

Main Methods:

• A statistical method employing a naive Bayes-like framework is proposed.
• The method combines three key quantities for each SNP: GWAS association p-value, cis-eQTL association p-value, and an rSNP prediction score.
• The rSNP prediction score can be derived from various sources, including machine-learning methods like CERENKOV2.

Main Results:

• The developed method effectively prioritizes candidate causal noncoding SNPs by integrating diverse genetic data.
• The approach is applicable to GWAS summary data, requiring only association p-values and not full genotype information.
• Demonstration using CERENKOV2 scores showcases the method's practical utility.

Conclusions:

• The proposed statistical method offers a robust approach for prioritizing noncoding SNPs in GWAS.
• Its reliance on summary data makes it a valuable tool for secondary GWAS analysis.
• This method enhances the ability to identify functional variants underlying trait associations.