Enabling improved applicability and transferability of polygenic scores across populations

About This Grant

Abstract The Polygenic Risk Methods Development (PRIMED) Consortium was established to harmonize large-scale genomic and phenotypic data and advance methodological innovation to improve polygenic risk scores (PRS). Within the scope of the PRIMED Consortium, the parent award: Functional and Fine-Mapping Approach to Improve Responsible Risk-modeling of Polygenic Risk Scores (FFAIRR-PRS, U01HG011719) aimed to harmonize and aggregate individual-level data across populations, to share analytic workflows and new statistical genetics frameworks, and to validate these approaches for enhanced polygenic score development, generalizability, and clinical translation. To that end, our team has developed new methods with requisite validation advancing new generalizable frameworks for polygenic risk scoring that have already translated to clinical adoption. Based on these new frameworks and consortium-wide learnings, the goal of this administrative supplement is to complete several ongoing high-impact consortium-wide and site-specific activities that will not yet be finalized within the current project period toward maximal impact for PRIMED. Our research team combines strengths in cardiovascular medicine, statistical genetics, and high throughput genetics and genomics. In Aim 1, we will combine existing PRS from the PGS Catalog using the PRSmix method to enhance cross-trait prediction accuracy and stability and will examine the scale of risk estimation and contextualization effects to optimize fairness in trait-specific PRS modeling across contexts. In Aim 2, we will finalize and disseminate new methodological frameworks, including integrative approaches that involve individual-specific trajectories and rare variant modeling. In Aim 3, we will extend benchmarking and integrative risk modeling efforts across key traits to improve clinical translation. Completion of our aims will further advance PRS accuracy and generalizability and enable methodological advances. We will integrate our learnings to both describe the current state of the field but also provide granular guidance regarding clinical translation and communication with patients.