A Prism Vote Framework for Individualized Risk Prediction of Traits in Genome-wide Sequencing Data of Multiple Populations

Multi-population cohorts offer unprecedented opportunities for profiling disease risk in large samples, however, heterogeneous risk effects underlying complex traits across populations make integrative prediction challenging. In this study, we propose a novel Bayesian probability framework, the Prism Vote (PV), to construct risk predictions in heterogeneous genetic data. The PV views the trait of an individual as a composite risk from subpopulations, in which stratum-specific predictors can be formed in data of more homogeneous genetic structure. Since each individual is represented by a composition of subpopulation memberships, the framework enables individualized risk characterization. Simulations demonstrated that the PV framework applied with alternative prediction methods significantly improved prediction accuracy in mixed and admixed populations. The advantage of PV enlarges as the sample size, genetic heterogeneity, and population diversity increase. In two real genome-wide association data consists of multiple populations, we showed that the framework enhanced prediction accuracy of the linear mixed model by up to 12.1% in five-group cross validations. The proposed framework offers a new aspect to analyze individual’s disease risk and improve accuracy for predicting complex traits in genome data.