Editorial Comment—Epistasis Is Coming

Abstract

For physicians and patients, the central challenge of human genetics is unraveling the genetic architecture of human diseases. Depending on the disease, this architecture can range from simple to complex. In the so-called mendelian diseases, 1 mutation in a single gene is sufficient to cause disease, whereas more complex diseases may involve changes in single genes, multiple genes, or even combinations of alleles within the same gene. The payoff for successful disease-gene discovery is likely to be enormous, leading to novel biologic discoveries and innovative interventions that can alleviate human suffering. The search so far, however, has been slower than we would all like. The nagging problem is that for most common diseases, the culprit gene variants contribute only a small proportion of disease risk. Successful studies must therefore detect relatively weak associations, and finding weak associations requires very large samples of patients. Epistasis, broadly defined, is the interaction between alleles at different loci, or positions, in the genome. Most simply, a variant at 1 locus can prevent a variant at another locus from manifesting its effect.1,2 These combinations of variants may be in different genes, or they may be within the same gene, eg, within the regulatory and coding regions. Ultimately, epistatic interaction is synergistic, with an effect that differs from the simple sum …