Leveraging pleiotropy to develop polygenic risk scores for cardiometabolic diseases - Abstract/Summary
Cardiometabolic diseases (CMD) are the leading causes of death worldwide. In this application we explore the
genomic risk for common CMD, including type 2 diabetes, coronary artery disease, hypertension, and
dyslipidemia, across non-European ancestry populations. Recent studies have shown that these complex human
traits are genetically correlated by pleiotropy, which occurs when a genetic locus affects more than one trait and
is a possible underlying cause for observed cross-phenotype associations. Relative to pleiotropy, polygenic risk
scores (PRS) which are estimated by combining GWAS-associated variants into a composite score, may better
explain the correlation between complex traits, by determining a subset of risk variants for each outcome.
However, PRS translate poorly when the discovery and target populations have differential ancestry. Combining
pleiotropy and PRS has great potential to uncover novel risk variants associated with CMD. We leverage the
large dataset in the All of Us research program and the NIH-funded CARdiometabolic Disorders IN African-
ancestry PopuLations (CARDINAL) study site, to identify pleiotropic variants and develop PRS for type 2
diabetes, coronary artery disease, hypertension, and dyslipidemia, in populations of diverse ancestry. The All of
Us dataset is ideal for generating novel biologic insights into complex disease etiology, with applications in global
populations.