Evolutionary footprint of epistasis

Epistasis is a widespread and ubiquitous genetic property of biological networks that shape evolutionary trajectories. Detecting epistasis in genomic data still represents a great challenge that could be met with a better understanding of epistasis from a mechanistic, evolutionary point of view. Here, using a standard model of population genetics applicable to viruses and microorganisms, we derive a universal relationship between four haplotype frequencies of a single pair in a genome, which depends only on the epistatic strength for the pair, defined regarding Darwinian fitness, but not on other system's parameters. Although we do not yet propose a tool for statistical inference of epistasis, we derive an analytic estimator of its strength from a single time DNA/RNA sequence database. The method is based on the balance existing between fitness and disorder caused by mutation and random genetic drift. Our technique can be used to understand the mechanism behind the rapid accumulation of mutation clusters that can cross the threshold of genetic stability and cause rapid adaptation of a virus to a new host, a new antiviral drug, or a new wave of immune response.