- Genetic correlation is a fundamental concept in quantitative genetics that describes the extent to which genetic effects influencing two traits are associated with one another. It measures whether individuals that have genetic effects increasing one trait also tend to have genetic effects increasing or decreasing another trait. Genetic correlation is particularly important for understanding complex traits, predicting correlated responses to selection, studying genetic architecture, and determining how selection on one trait may influence another.
- Genetic correlation is closely related to genetic covariance and genetic variance. Genetic covariance measures the direction and magnitude of the joint variation in genetic effects between two traits, while genetic correlation standardizes genetic covariance by the genetic standard deviations of the two traits. It can be expressed as rG=CovG(A1,A2)VA1VA2r_G = \frac{Cov_G(A_1,A_2)}{\sqrt{V_{A1}V_{A2}}}, where CovG(A1,A2)Cov_G(A_1,A_2) is the additive genetic covariance between traits and VA1V_{A1} and VA2V_{A2} are their additive genetic variances. The resulting correlation ranges from −1 to +1.
- A positive genetic correlation means that genetic effects associated with higher values of one trait tend to be associated with higher values of the other trait. For example, growth rate and mature body size may have a positive genetic correlation in some populations. A negative genetic correlation means that genetic effects increasing one trait tend to be associated with lower values of another trait. A correlation close to zero indicates little linear association between the additive genetic effects influencing the two traits under the conditions studied.
- The magnitude of a genetic correlation indicates the strength of the association. A value near +1 or −1 indicates a strong genetic association, whereas a value near zero indicates a weak linear genetic association. However, a genetic correlation of zero does not necessarily mean that the two traits have no biological relationship. Nonlinear relationships, interactions, or different forms of genetic association may exist without producing a substantial linear genetic correlation.
- Genetic correlation should be distinguished from phenotypic correlation. Phenotypic correlation measures the association between observed trait values, whereas genetic correlation measures the association between genetic effects underlying those traits. Two traits may have a strong phenotypic correlation because of shared environmental conditions while having a weak genetic correlation. Conversely, traits may have a modest phenotypic correlation but a strong genetic correlation if environmental effects obscure their underlying genetic association.
- Environmental covariance can therefore influence phenotypic relationships without necessarily producing genetic correlation. Individuals may share environmental conditions, developmental environments, management practices, nutrition, or other external influences that cause two traits to vary together. Separating genetic and environmental sources of covariance is an important part of multivariate quantitative-genetic analysis.
- One of the major biological causes of genetic correlation is pleiotropy, in which a single genetic variant or gene influences multiple traits. If the same genetic variants affect two traits, changes in those variants can produce coordinated changes in both traits. Pleiotropy can therefore create positive or negative genetic correlations depending on the direction and magnitude of the effects on each trait.
- Another potential source of genetic correlation is linkage disequilibrium (LD), which occurs when alleles at different genetic loci are associated with one another more frequently than expected under random association. If variants influencing two traits are located at linked loci, nonrandom associations between alleles can create genetic covariance and therefore genetic correlation. The contribution of LD can change as recombination, selection, migration, drift, and population structure alter allele frequencies and haplotype associations.
- Genetic correlation is therefore influenced by genetic architecture. Traits controlled by many genes may share genetic variants, biological pathways, regulatory mechanisms, or linked genomic regions. The architecture of the traits determines whether their genetic effects tend to be aligned, opposed, or largely independent. Polygenic inheritance, pleiotropy, epistasis, dominance, and LD can all contribute to the genetic relationships observed between traits.
- Genetic correlation is especially important in multivariate quantitative genetics, where several traits are studied simultaneously. Rather than analyzing each trait independently, researchers can estimate a matrix of additive genetic variances and covariances known as the G-matrix. The diagonal elements of the G-matrix represent additive genetic variances for individual traits, while the off-diagonal elements represent additive genetic covariances between traits. Genetic correlations can be derived from these variance and covariance components.
- The G-matrix provides a framework for predicting how populations respond to selection on multiple traits. The multivariate breeder’s equation can be expressed as Δz=Gβ\Delta \mathbf{z} = \mathbf{G}\boldsymbol{\beta}, where Δz\Delta \mathbf{z} represents the vector of predicted changes in trait means, G\mathbf{G} is the additive genetic variance-covariance matrix, and β\boldsymbol{\beta} represents the vector of selection gradients. Genetic correlations within the G-matrix can therefore influence the direction and magnitude of evolutionary or breeding responses.
- A particularly important consequence of genetic correlation is correlated response to selection. When selection is applied to one trait, genetic changes may occur in another trait because the traits share additive genetic effects. For example, selecting animals for increased growth may unintentionally change body composition, reproductive traits, feed efficiency, or disease resistance if those traits are genetically correlated with growth.
- Correlated responses can be favorable or unfavorable. A favorable genetic correlation can allow simultaneous improvement in multiple traits. An unfavorable correlation can create a genetic trade-off, where improvement in one trait is accompanied by deterioration in another. Such trade-offs are common in breeding and evolutionary biology and can constrain the direction of long-term selection.
- Genetic correlations are particularly important in animal breeding and plant breeding. Breeders rarely select for a single trait in isolation. Agricultural programs may simultaneously consider yield, growth, fertility, disease resistance, product quality, environmental adaptation, and other characteristics. Genetic correlations help breeders predict whether improving one trait will produce desirable or undesirable changes in others.
- The use of selection indices is one practical application of genetic correlations. Selection indices combine information from multiple traits to identify individuals with the most favorable overall genetic potential. Estimates of genetic variance and covariance are essential for constructing these indices and determining appropriate economic or biological weights for different traits.
- Genetic correlation also influences the accuracy of breeding value prediction. When multiple genetically correlated traits are measured, information from one trait can provide information about the genetic value of another. This can be particularly useful when one trait is difficult, expensive, late in life, or impossible to measure directly in selection candidates.
- This principle is important in genomic selection. Genome-wide genetic markers can be used to estimate relationships among individuals and predict genomic estimated breeding values (GEBVs) for multiple traits. Multivariate genomic prediction models can exploit genetic correlations among traits to improve prediction accuracy, especially when some traits have limited phenotypic records.
- Genetic correlation is also closely related to heritability. Heritability describes the proportion of phenotypic variance associated with additive genetic variance for a single trait, while genetic correlation describes the association between additive genetic effects for two traits. Two traits can both have high heritability but a low genetic correlation, or both have moderate heritability and a strong genetic correlation.
- The magnitude of a genetic correlation depends partly on the amount of additive genetic variance in each trait. Because the correlation is standardized by the genetic standard deviations of the traits, it describes the relative association between their genetic effects rather than their absolute scale. This allows traits measured in different units to be compared genetically.
- Genetic correlations can be estimated using covariance among relatives. Family members share predictable proportions of their genomes, so the covariance between relatives for multiple traits can provide information about additive genetic covariance. Parent-offspring, full-sib, half-sib, and other relationship structures can therefore be used to estimate genetic correlations.
- Modern methods can also estimate genetic correlations using genomic data. A genomic relationship matrix constructed from genome-wide markers can be incorporated into mixed models to estimate additive genetic variances and covariances. These approaches can provide more detailed estimates than traditional pedigree-based methods, particularly when large genomic datasets are available.
- Statistical methods such as linear mixed models and restricted maximum likelihood (REML) are commonly used to estimate genetic covariance and correlation. These models can account for fixed effects, environmental effects, repeated measurements, pedigree relationships, genomic relationships, and other sources of variation. Because estimates are subject to sampling uncertainty, confidence intervals or standard errors are important when interpreting genetic correlations.
- Genetic correlation can also vary among populations. Differences in allele frequencies, LD patterns, genetic diversity, selection history, demographic history, and genetic architecture can alter the covariance structure between traits. A genetic correlation estimated in one population should therefore not automatically be assumed to apply to another population.
- Environmental conditions can further modify genetic relationships through genotype–environment interaction (G×E). Genetic effects may differ across environments, causing genetic correlations between traits to vary among locations, climates, diets, management systems, or developmental conditions. Multienvironment breeding programs therefore often evaluate genetic correlations across environments to determine whether selection will produce consistent responses.
- Genetic correlation has an important role in evolutionary genetics because it can create evolutionary constraints. If two traits are strongly and negatively genetically correlated, natural selection favoring an increase in both traits may face a genetic trade-off. Conversely, positive genetic correlations can facilitate coordinated evolutionary change. The structure of genetic covariance can therefore influence the direction and rate of adaptation.
- The concept is also important in the study of life-history evolution. Traits such as growth, reproduction, survival, development, and resource allocation may be genetically correlated because they share biological pathways or pleiotropic genetic effects. These correlations can create trade-offs that influence evolutionary strategies and population responses to natural selection.
- Genetic correlations can change over time. Mutation introduces new genetic variation, recombination reshuffles genetic combinations, natural selection changes allele frequencies, genetic drift alters variation randomly, and gene flow introduces genetic material between populations. These processes can modify the genetic covariance structure and consequently alter genetic correlations across generations.
- Genetic correlation is also relevant to QTL mapping and GWAS. If the same genomic regions are associated with multiple traits, this may provide evidence for pleiotropy or closely linked causal variants. Multitrait genetic analyses can therefore help distinguish shared genetic architecture from relationships produced primarily by environmental covariance.
- However, a genetic correlation does not by itself demonstrate that one trait causes another. A strong genetic correlation may result from pleiotropy, LD, shared biological pathways, population structure, or other features of genetic architecture. Understanding the biological mechanism behind a genetic correlation requires additional genetic, molecular, experimental, or functional evidence.
- A genetic correlation should also not be interpreted as a permanent or universal biological constant. It is a statistical property of a particular population, environment, and set of genetic conditions. Its value may change when allele frequencies, environmental conditions, selection pressures, or genetic architecture change. This population-specific nature is especially important when applying estimates from research populations to breeding or natural populations.
- In human genetics, genetic correlations can be used to investigate shared genetic influences among complex traits and diseases. Large-scale genomic datasets can reveal genetic overlap between traits that may have apparently different biological manifestations. However, interpreting such correlations requires careful consideration of population structure, LD, pleiotropy, ascertainment, and the statistical methods used to estimate them.
- Overall, genetic correlation provides a quantitative measure of how genetic effects influencing two traits vary together. It is derived from genetic covariance and genetic variance and plays a central role in understanding pleiotropy, linkage disequilibrium, genetic architecture, correlated response to selection, and multivariate evolution.
- Genetic correlation is essential for predicting correlated responses to selection, designing multi-trait breeding programs, estimating genetic relationships among complex traits, understanding evolutionary constraints, and interpreting the G-matrix. When combined with heritability, breeding value, genetic covariance, genomic selection, and information from QTL and GWAS, it provides a powerful framework for studying the inheritance and evolution of multiple traits simultaneously.
