Enhancing polygenic scores for cardiometabolic traits through tissue- and cell-type-specific functional annotations

Area of Science:

• Genomics
• Cardiovascular Genetics
• Metabolic Disease Research

Background:

• Polygenic scores (PGS) predict complex traits using genome-wide association study (GWAS) data.
• General genomic annotations improve PGS, but tissue- and cell-type-specific annotations may offer further enhancement for cardiometabolic traits.

Purpose of the Study:

• To evaluate the impact of functional genomic annotations, including tissue-specific data, on polygenic score development for 14 cardiometabolic traits.
• To compare the performance of PGS built with different annotation strategies and variant sets.

Main Methods:

• Developed polygenic scores (PGS) for 14 cardiometabolic traits using SBayesRC in the UK Biobank.
• Integrated GWAS summary statistics with general annotations (Baseline-LD), cell-type-specific snATAC-seq peaks, and tissue-specific eQTLs/sQTLs.
• Utilized two European (EUR) LD reference panels: HapMap3 (1.2M variants) and 7M imputed variants.

Main Results:

• Tissue- and cell-type-specific annotations demonstrated stronger heritability enrichment than general annotations.
• The 7M variant PGS significantly outperformed the HapMap3 variant PGS when using all annotations (8% average increase in EUR).
• All annotation strategies improved PGS performance compared to no annotations, with general and all annotations yielding the largest gains for the 7M variant set.

Conclusions:

• Functional genomic annotations effectively improve polygenic scores for cardiometabolic traits.
• While tissue- and cell-type-specific annotations show high heritability enrichment, their performance gains over general annotations were marginal.
• Annotations improved cross-ancestry prediction but did not reduce performance disparities between genetic ancestry groups.