A proteome-wide association study of cardiovascular diseases in 640,000 participants of multiple ancestries

The authors developed genetic imputation models for 2,594 plasma proteins using data from 54,219 UK Biobank participants, then validated these models across ancestry groups and an independent cohort before imputing proteomes for 640,000+ individuals. This allows testing of protein-disease associations at population scale without direct proteomic measurements. The focus on cardiovascular diseases covers conditions such as coronary artery disease, stroke, and heart failure. By leveraging multi-ancestry data, the study aims to improve generalizability and uncover ancestry-specific signals that may be missed in European-only studies. The findings could prioritize druggable proteins and inform future drug development pipelines.