Flexibly encoded GWAS identifies novel nonadditive SNPs in individuals of African and European ancestry

Abstract

AbstractMost genome-wide association studies (GWASs) assume an additive inheritance model, where heterozygous genotypes (HET) are coded with half the risk of homozygous alternate genotypes (HA), leading to less explained nonadditive genetic effects for complex diseases. Yet, growing evidence indicates that with flexible modeling, many single-nucleotide polymorphisms (SNPs) show nonadditive effects, including dominant and recessive, which will be missed using only the additive model. We developed Elastic Data-Driven Encoding (EDGE) to determine the HET to HA ratio of risk. Simulation results demonstrated that EDGE outperformed traditional methods across all simulated models for power while maintaining a conserved false positive rate. This research lays the necessary groundwork for integrating nonadditive genetic effects into GWAS workflows to identify novel disease-risk SNPs, which may ultimately improve polygenic risk prediction in diverse populations and springboard future applications to thousands of disease phenotypes and other omic domains to improve disease-prediction capability.