Two variance component model improves genetic prediction in family data sets

Abstract Genetic prediction based on either identity by state (IBS) sharing or pedigree information has been investigated extensively using Best Linear Unbiased Prediction (BLUP) methods. Such methods were pioneered in the plant and animal breeding literature and have since been applied to predict human traits with the aim of eventual clinical utility. However, methods to combine IBS sharing and pedigree information for genetic prediction in humans have not been explored. We introduce a two variance component model for genetic prediction: one component for IBS sharing and one for approximate pedigree structure, both estimated using genetic markers. In simulations using real genotypes from CARe and FHS family cohorts, we demonstrate that the two variance component model achieves gains in prediction r 2 over standard BLUP at current sample sizes, and we project based on simulations that these gains will continue to hold at larger sample sizes. Accordingly, in analyses of four quantitative phenotypes from CARe and two quantitative phenotypes from FHS, the two variance component model significantly improves prediction r 2 in each case, with up to a 20% relative improvement. We also find that standard mixed model association tests can produce inflated test statistics in data sets with related individuals, whereas the two variance component model corrects for inflation. Author Summary Genetic prediction has been well-studied in plant and animal breeding and has generated considerable recent interest in human genetics, both in family data sets and in population cohorts. Many prediction studies are based on the widely used Best Linear Unbiased Prediction (BLUP) approach, which performs a mixed model analysis using a genetic relationship matrix that is either estimated from genotype data—thus measuring identity-by-state (IBS) sharing—or obtained from family pedigree information. We show here that a substantial improvement in prediction accuracy in family data sets can be obtained by jointly modeling both IBS sharing and approximate pedigree structure, both estimated using genetic markers, using separate variance components within a two variance component mixed model. We demonstrate the performance of this model in simulations and real data sets. We also show that previous mixed model association methods suffer from inflated test statistics in family data sets due to their failure to account for the different heritability parameters corresponding to IBS sharing vs. pedigree relatedness. Our two variance component model provides a solution to this problem without compromising statistical power.