P119: Toward best-practice validation of disease polygenic scores for adult screening: Within-family, liability-scale, and ancestry-aware methods*

Polygenic risk scores (PRS) are increasingly proposed for clinical risk stratification and preventive screening, yet methodological standards for their validation and interpretation remain inconsistent. We present an integrated framework for “fit-for-clinic” validation of disease PRS that unifies within-family evaluation, liability-scale variance assessment for non-normal predictors, and ancestry-aware calibration of effect sizes. Building on recent large-scale disease PRS analyses, we leverage snipar-based parental genotype imputation to perform within-family regression of disease outcomes on offspring PRS, providing direct-effect estimates that are robust to population stratification, assortative mating, and indirect genetic effects.