While the trans-ancestry PRS derived by PRS-CSx showed promise for clinical translation when evaluated separately in each population, implementation of the PRS in a prospective cohort with individuals from diverse ancestry backgrounds requires calibrated PRS distributions across populations. In this work, we evaluated a post hoc ancestry adjustment method that can express the polygenic risk on the same scale across ancestries without compromising the discrimination capability at the extreme tails of the PRS. With this adjustment, a single cutoff of the PRS distribution for the high-risk group can be identified and applied to any target individual. We expect that using large and diverse reference panels that better match the population structure of the prospective cohort will facilitate more accurate ancestry adjustment and risk estimation.
