The contribution of gametic phase disequilibrium to the heritability of complex traits.

Nonrandom mating induces genome-wide correlations between unlinked genetic variants, known as gametic phase disequilibrium (GPD), whose contribution to heritability remains uncharacterized. Here we introduce the disequilibrium genome-based restricted maximum likelihood (DGREML) method to simultaneously quantify the additive contribution of SNPs to heritability and that of their directional covariances. We applied DGREML to 26 phenotypes of 550,000 individuals from diverse biobanks and found that cross-autosome GPD contributes 10-27% of the SNP-based heritability of height, educational attainment, intelligence, income, self-rated health status and sedentary behaviors. We observed a differential contribution of GPD to the heritability of height between the UK, Chinese and Japanese populations. Finally, bivariate DGREML analyses of educational attainment and height show that cross-autosome GPD contributes at least 32% of their genetic correlation. Altogether, our versatile and powerful method reveals understudied features of the genetic architecture of complex traits and informs potential mechanisms generating these features.