The inheritance of traits generally follows independent assortment, where genetic variants at different locations (loci) are passed down randomly. This random process ensures that the combination of alleles received for different genes is mostly unpredictable. However, this expectation of independence often breaks down when considering the alleles present across a large group of individuals. Linkage disequilibrium (LD) describes a situation where alleles at two or more loci are associated with each other more often than expected by chance. This non-random association means observing a specific variant at one location makes it more likely that a particular variant will also be observed at a nearby location. LD is a property of a population, representing a deviation from the expected random mixing of genetic variants over time.
Understanding Linkage Disequilibrium
The concept of Linkage Disequilibrium is frequently confused with genetic linkage, but they are distinct phenomena. Genetic linkage refers to the physical state where two loci are located close together on the same chromosome. This physical proximity limits the chance of separation during inheritance, meaning close loci are generally inherited together.
Linkage disequilibrium, by contrast, is a statistical measure of association between alleles in a population, regardless of their physical distance. For instance, two physically linked alleles might be in “linkage equilibrium” if their combination has been broken apart by many generations of genetic shuffling, resulting in a random association. Conversely, two distant alleles could be in LD due to specific population events.
The non-random association of alleles in LD results in the formation of a haplotype, which is a set of specific alleles found together on the same chromosome. These haplotypes are blocks of genetic information passed down as a unit. When two variants are in strong LD, they are part of the same co-inherited block of DNA, making the combination of their alleles highly predictable within that population.
Quantifying Linkage Disequilibrium
Scientists use specific statistical metrics to measure the strength of this non-random association between alleles. The two most common measures are $D’$ (D-prime) and $r^2$ (R-squared), which quantify the degree of LD between a pair of genetic variants. These metrics are calculated by comparing the observed frequencies of the four possible two-locus haplotypes to the frequencies expected under linkage equilibrium.
The $D’$ metric measures the deviation from random association, indicating the amount of historical recombination that has occurred between the two sites. A $D’$ value close to 1 suggests the two alleles have never been separated by recombination since the haplotype originated. This metric is useful for understanding historical constraints on recombination within a region.
The $r^2$ value ranges from 0 to 1 and measures the statistical correlation between the two loci. An $r^2$ value of 1 means one genetic variant perfectly predicts the other, making them redundant for statistical studies. Because $r^2$ directly reflects how much information one variant provides about the other, it is the preferred metric for applications like gene mapping, where a higher value indicates stronger statistical utility.
Biological Forces Driving Linkage Disequilibrium
The patterns of linkage disequilibrium across the genome are shaped by evolutionary and demographic forces. The primary force that continuously breaks down LD is recombination, the process of genetic material exchange during the formation of reproductive cells. The closer two genetic variants are on a chromosome, the less likely recombination is to occur between them, allowing LD to persist for more generations.
Biological processes and population history can also create or maintain LD over long genetic distances. Positive natural selection, for instance, can cause a beneficial mutation to rise rapidly in frequency in a “selective sweep.” As the new variant spreads, it carries a large block of surrounding genetic material, creating a region of strong LD that persists until recombination breaks it apart.
Population-level events also impact LD patterns across the genome. Genetic drift, the random fluctuation of allele frequencies due to small population size or bottlenecks, can randomly fix certain allele combinations, thereby creating LD. Furthermore, the recent mixing of distinct populations, known as admixture, can introduce new haplotypes and inflate LD by bringing together variants common in different ancestral groups.
Using LD for Genetic Discovery
Linkage disequilibrium enables the efficient mapping of genes associated with complex traits and diseases. A key application is in Genome-Wide Association Studies (GWAS), which scan the entire genome to find variants statistically associated with a particular trait, such as disease susceptibility. LD makes it possible to perform these studies without sequencing every single base pair.
This efficiency is achieved through indirect association, often called “tagging.” Researchers do not need to directly test a disease-causing mutation if it is in strong LD with a common, easily genotyped marker variant. The marker, or “tag SNP,” acts as a proxy. Its strong correlation ($r^2$) with the causative variant ensures that the statistical association found at the marker location accurately points to the true causal region.
By capitalizing on LD blocks—regions where haplotypes have been preserved—scientists significantly reduce the number of genetic markers they need to examine. This strategy has been instrumental in identifying thousands of genetic loci linked to human health and disease. Furthermore, the characteristic patterns of LD decay across the genome provide a molecular clock that allows researchers to reconstruct the historical size and migration patterns of human populations.