Linkage Disequilibrium

AbstractWhen two or more polymorphic loci are studied in a population, the interaction between the loci is often expressed in terms of linkage disequilibrium (LD). The loci are in LD if their respective alleles do not associate independently (randomly). LD does not necessarily imply physical linkage, however most often the loci considered are on the same chromosome and the degree of over‐ or underrepresentation of an expected haplotype measures the extent of LD between a specific pair of alleles. Multilocus patterns of LD are often visualised graphically, revealing local blocks of high LD. LD is generated by mutation but may also be generated and maintained by population processes including selection, drift and admixture. Genomic rearrangements, such as inversions may also influence LD patterns. Recombination overtime acts to reduce LD and this relationship may be used to date alleles.Key Concepts:Linkage disequilibrium described the nonrandom association between pairs of alleles at different loci.The loci under consideration are typically, but not necessarily, physically linked (on the same chromosome) and the allelic interactions can be described in terms of haplotype or ‘gamete’ frequencies and their deviation from expectation under independence.LD coefficients refer to specific pairs of alleles and are often normalised to minimise their dependence on allele frequencies. Higher‐order multiallelic or multiple‐loci coefficients exist but are complex.Multiple, linked loci typically exhibit blocks of high LD interspersed by areas of higher recombination.De novomutation generates perfect LD between the mutated site and flanking polymorphic sites. This LD is eroded overtime by recombination, the rate of decay being function of physical distance between the loci. This relationship can be used to date alleles.The presence of four haplotypes (orD′<1) in a two‐locus bi‐allelic system (e.g. a pair of SNPs) implies that at least one recombination event has occurred.LD may be generated and maintained by positive and balancing selection, genetic drift in small populations, admixture and gene flow and genomic rearrangements.LD forms the basis for mapping complex disease association by association.