Identifying causal variants by fine mapping across multiple studies.

Area of Science:

• Genetics and Genomics
• Statistical Genetics
• Bioinformatics

Background:

• Large-scale Genome-Wide Association Studies (GWAS) identify numerous variants linked to complex traits.
• Fine-mapping methods are crucial for pinpointing causal variants among associated ones.
• Availability of GWAS from diverse populations offers opportunities to refine fine-mapping by utilizing varying linkage disequilibrium (LD) structures.

Purpose of the Study:

• To develop a novel method, MsCAVIAR, for identifying causal variants in associated regions across multiple GWAS.
• To extend the CAVIAR fine-mapping framework to a multi-study setting.
• To leverage diverse LD structures from different populations for improved fine-mapping accuracy.

Main Methods:

• Introduction of multiple study causal variants identification in associated regions (MsCAVIAR), a method using a random effects model.
• MsCAVIAR requires only summary statistics and LD information as input.
• The method employs a multivariate normal model to handle uncertainty in association statistics and models varying SNP effect sizes across populations, allowing for multiple causal variants per locus.

Main Results:

• Demonstrated the effectiveness of MsCAVIAR through a simulation study.
• Validated MsCAVIAR in a trans-ethnic, trans-biobank fine-mapping analysis of High Density Lipoprotein (HDL) levels.
• The method successfully refines fine-mapping results by integrating information from multiple studies with different LD patterns.

Conclusions:

• MsCAVIAR provides an effective framework for multi-study fine-mapping of causal variants.
• The method enhances the identification of causal variants by accounting for population-specific LD structures and varying effect sizes.
• MsCAVIAR represents a significant advancement in leveraging diverse GWAS data for precise genetic variant discovery.