Integrated multiomic scores for cardiovascular disease prediction across multiancestral populations

Each standard deviation increase in the corresponding MEGASCORE in PURE is linked to higher risk of heart failure, with HR 3.33 (95% CI, 2.75–4.05).

• Why it matters: Genomic, epigenomic, and proteomic signals may complement each other, but it is uncertain whether integrated multiomic prediction works across ancestry groups for multiple cardiovascular outcomes.
• What they did: Disease-specific proteomic scores from 999 plasma proteins were derived in 50,673 UK Biobank participants and validated in 9,358 PURE participants, then integrated with 2,279 polygenic and 479 DNA methylation scores to form four MEGASCOREs.
• The result: MEGASCORE-CV-Death improved cardiovascular mortality discrimination versus SCORE2 (AUC 0.83 vs 0.78) and showed higher discrimination across all four ancestry groups with better calibration in non-European groups.