CoPheScan: phenome-wide association studies accounting for linkage disequilibrium

Area of Science:

• Genetics and Genomics
• Statistical Genetics
• Bioinformatics

Background:

• Phenome-wide association studies (PheWAS) link genetic variants to multiple traits, aiding drug discovery and side effect identification.
• Traditional PheWAS struggles to differentiate true pleiotropy from confounding due to linkage disequilibrium (LD).

Purpose of the Study:

• To introduce CoPheScan (Coloc adapted Phenome-wide Scan), a Bayesian method to address LD confounding in PheWAS.
• To systematically explore causal genetic associations while controlling for LD.

Main Methods:

• Developed CoPheScan, a Bayesian approach adapting COLOC for genome-wide scans.
• Validated CoPheScan through simulations, comparing its false positive rate control against conventional methods.
• Applied CoPheScan to UK Biobank data, analyzing protein-truncating and fine-mapped variants across 2275 disease phenotypes.

Main Results:

• CoPheScan demonstrated superior control of false positive rates compared to approaches ignoring LD.
• Analysis revealed the complex pleiotropic effects of known genes like APOE.
• Identified a potential causal role for TGM3 in the development of skin cancer.

Conclusions:

• CoPheScan provides a robust framework for accurate causal inference in PheWAS.
• The method enhances the ability to uncover true pleiotropic effects and novel genotype-phenotype relationships.
• Findings suggest TGM3 as a potential target for skin cancer research.