Characterization of shared and ancestry-specific signals driving complex traits using multi-ancestry fine-mapping

Area of Science:

• Population Genetics
• Statistical Genomics
• Complex Trait Analysis

Background:

• Genome-wide association studies (GWAS) often identify shared genetic signals across populations, but ancestry-specific signals also exist.
• Fine-mapping methods struggle to distinguish causal variants in high linkage disequilibrium (LD) and to study ancestry-specific contributions.
• Existing multi-study fine-mapping methods often assume shared causal variants, limiting the study of population-specific effects.

Purpose of the Study:

• To introduce PIPSORT, a novel multi-study fine-mapping method for simultaneously detecting shared and ancestry-specific genetic signals.
• To quantify the evidence of signal sharing across different ancestral populations.
• To investigate the genetic architecture of platelet count, LDL cholesterol, and schizophrenia across diverse ancestries.

Main Methods:

• Developed PIPSORT, a multi-study fine-mapping framework designed to detect both shared and ancestry-specific signals.
• Applied PIPSORT to fine-map platelet count and LDL cholesterol in African and European ancestry groups within the UK Biobank and All of Us datasets.
• Extended the analysis to fine-map schizophrenia signals in East Asian versus European populations and explored ancestry interactions within admixed populations.

Main Results:

• PIPSORT successfully identified numerous ancestry-specific signals for platelet count and LDL cholesterol, with 10 novel signals detected in the diverse All of Us cohort.
• Known African-specific variants in MPL and PCSK9 were confidently identified, highlighting PIPSORT's ability to detect population-specific effects.
• Analysis of schizophrenia in East Asians revealed multiple ancestry-specific signals, including a variant in SLC39A8. Admixture analysis in All of Us identified ancestry-by-ancestry interaction signals.

Conclusions:

• The study demonstrates that multiple strong ancestry-specific signals exist for complex traits, offering new insights into their genetic architecture across diverse populations.
• PIPSORT effectively distinguishes shared and ancestry-specific signals, outperforming previous methods in resolution and applicability.
• These findings have significant implications for developing future multi-ancestry methods for complex trait analysis and personalized medicine.