Understanding the genetic risks of complex diseases using the additive epistatic interaction model: a simulation study

Abstract Thousands of Genome-Wide Association Studies (GWAS) have been carried out to pinpoint genetic variants associated with complex diseases. However, the proportion of phenotypic variance which can be explained by the identified genetic variants is relatively low, leading to the “missing heritability” problem. This problem may be partly caused by the inadequate understanding of the genetic mechanisms of complex diseases. Here, we propose the additive epistatic interaction model, consisting of widespread pure epistatic interactions whose effects are additive and can be summarized by a genetic risk score. Based on a simulated genotype dataset, the additive epistatic interaction model well depicted genetic risks and hereditary patterns of complex diseases. Based on the 1000 Genomes Project data, the additive epistatic interaction model accurately classified human populations. Moreover, the model’s genetic risk score can be replaced by a deep learning model which is more resistant to noises. We suggest that the additive epistatic interaction model may help to understand the genetic mechanisms and risks of complex diseases.