Functionally informed fine-mapping and polygenic localization of complex trait heritability.

Area of Science:

• Genetics
• Genomics
• Computational Biology

Background:

• Fine-mapping seeks to pinpoint causal genetic variants influencing complex traits.
• Current methods often focus on genome-wide significant loci, potentially missing important variants.

Purpose of the Study:

• To introduce PolyFun, a scalable framework for improving genetic fine-mapping accuracy.
• To leverage functional annotations across the entire genome to inform fine-mapping priors.
• To enhance the identification of causal variants for complex traits.

Main Methods:

• PolyFun integrates functional annotations genome-wide to set prior probabilities for fine-mapping tools (e.g., SuSiE, FINEMAP).
• The framework was evaluated using simulations and analyses of 49 UK Biobank traits (average n=318,000).
• Posterior mean per-SNP heritabilities were used for polygenic localization.

Main Results:

• PolyFun combined with SuSiE/FINEMAP identified >20% more variants with posterior causal probability >0.95 in simulations compared to non-functionally informed methods.
• In UK Biobank data, PolyFun+SuSiE identified 3,025 fine-mapped variant-trait pairs (>32% improvement over SuSiE alone).
• PolyFun enabled polygenic localization, identifying minimal sets of SNPs explaining 50% of heritability for various traits.

Conclusions:

• PolyFun significantly improves fine-mapping accuracy by utilizing comprehensive functional genomic data.
• The framework effectively prioritizes variants for functional follow-up studies.
• PolyFun offers novel insights into the genetic architecture of complex human traits.