Epistatic redundancy reconciles pervasive genetic interactions with their rarity in genome-wide analyses.

Abstract

Epistasis, the genetic interaction between loci, has long been recognized as a fundamental feature of the genetic architecture, but its importance remains controversial. This controversy stems partly from the apparent disconnect between abundant physiological epistasis observed in laboratory settings and the rarity of detectable epistatic interactions in genome-wide association studies (GWAS). Here, I introduce the concept of epistatic redundancy, whereby a focal locus interacts with multiple alternative loci rather than a single partner. Using population genetic simulations, I demonstrate that epistatic redundancy substantially amplifies the influence of epistasis compared to models assuming only pairwise interactions, particularly under U-shaped allele frequency distributions, which are characteristic of natural populations. This framework reconciles theoretical predictions with empirical observations and suggests that current epistasis detection approaches, which focus on pairwise interactions, systematically underestimate the contribution of epistasis. I discuss alternative approaches to detect epistatic redundancy in both association studies and evolving populations.