JAM: A Scalable Bayesian Framework for Joint Analysis of Marginal SNP Effects

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Large-scale genome-wide association studies (GWAS) identify numerous trait-associated genetic variants.
• Current analysis methods typically examine variants individually, hindering precise fine-mapping.
• Identifying a refined set of single nucleotide polymorphisms (SNPs) for functional studies is challenging.

Purpose of the Study:

• To introduce a scalable algorithm, Joint Analysis of Marginal Summary Statistics (JAM), for analyzing multiple SNPs simultaneously.
• To improve the fine-mapping process by re-analyzing published marginal summary statistics under joint multi-SNP models.
• To enhance the identification of causal variants from complex genomic regions.

Main Methods:

• Developed a scalable algorithm, JAM, for joint multi-SNP analysis using published marginal summary statistics.
• Accounted for SNP correlation using estimates from a reference dataset.
• Employed an integrated Bayesian penalized regression framework to identify significant joint patterns of marginal effects.
• Implemented JAM using enumerated and Reversible Jump Markov Chain Monte Carlo (MCMC) approaches.

Main Results:

• JAM demonstrated comparable performance to single-SNP methods in single-region settings.
• In multi-region settings, JAM showed superior power and specificity compared to stepwise selection methods.
• Application to the Meta-Analysis of Glucose and Insulin-related Traits Consortium (MAGIC) data identified potential additional SNPs for follow-up, including a biologically plausible SNP in the ADCY5 gene.

Conclusions:

• JAM provides a powerful and scalable approach for joint multi-SNP analysis of GWAS data.
• The algorithm effectively refines SNP sets for functional follow-up, particularly in regions with multiple signals.
• JAM enhances the discovery of biologically relevant variants by enabling joint multivariate modeling.