Multi-resolution localization of causal variants across the genome

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) face challenges in pinpointing causal variants due to linkage disequilibrium.
• Maximizing statistical power while minimizing false positives is crucial for accurate genetic analysis.

Purpose of the Study:

• To introduce KnockoffZoom, a novel method for precise localization of causal variants in GWAS.
• To develop a flexible approach that controls the false discovery rate (FDR) at multiple resolutions.

Main Methods:

• KnockoffZoom tests conditional associations of genetic segments with decreasing width.
• The method employs artificial genotypes as negative controls.
• It is applicable to both quantitative and binary phenotypes without genetic architecture assumptions, relying on linkage disequilibrium models.

Main Results:

• KnockoffZoom achieves fine-mapping precision and detects more associations than traditional mixed-effects models.
• The method provably controls the false discovery rate.
• Application to UK Biobank data (350k subjects) revealed numerous novel findings.

Conclusions:

• KnockoffZoom offers a powerful and flexible tool for fine-mapping causal variants in large-scale genetic studies.
• The method enhances the accuracy of genetic association analysis and discovery of trait-associated variants.