Epistasis: genetic interactions between loci

Types and classical examples

Geneticists recognize several patterns of epistasis. In recessive epistasis one recessive allele at one locus masks alleles at another; in dominant epistasis a dominant allele masks effects elsewhere. Duplicate genes can produce redundancy so that mutation of one locus has no visible effect unless the other is also altered. Famous textbook examples include pigmentation pathways, such as coat color in dogs and mice, where one gene controls pigment production and another controls pigment deposition. When one locus prevents pigment formation entirely, it makes the state of pigmentation genes irrelevant, a clear epistatic relationship.

Biological versus statistical epistasis

Two complementary meanings of epistasis are commonly used. Biological (molecular) epistasis refers to mechanistic interactions among gene products or biochemical steps—how proteins and pathways influence one another. Statistical epistasis describes departures from additivity detected in genetic experiments or association studies: it is the interaction term in a model that cannot be explained by single-locus effects alone. A molecular interaction will often produce statistical epistasis, but complex networks, environmental effects, and allele frequencies all shape whether interactions are detectable in population data.

Importance, detection and implications

Epistasis matters for breeding, disease genetics and evolution. It can obscure the mapping of trait-causing loci, complicate prediction of hybrid performance (heterosis), and influence how populations respond to selection. Detecting epistasis requires experiments or statistical models that include interaction terms, such as controlled crosses, quantitative trait locus (QTL) mapping, or genome-wide association studies that look for pairs or higher-order combinations of loci. Model organisms—fruit flies, mice, yeast and plants—have provided many instructive examples where controlled crosses reveal epistatic architectures.

• Practical consequence: epistasis can make phenotype prediction from genotype more difficult than single-gene models imply.
• Evolutionary consequence: epistasis contributes to rugged fitness landscapes and can promote or hinder adaptation depending on genetic context.
• Research note: quantitative geneticists separate additive variance from epistatic (interaction) variance when partitioning trait heritability.

Understanding epistasis requires combining classical genetic crosses, molecular biology of pathways, and modern statistical genetics. For readers seeking historical context and foundational experiments, sources linked in the text offer entry points to the original work and later developments in quantitative genetics and evolutionary theory.