The use of linkage disequilibrium to map quantitative trait loci
This paper reviews the causes of linkage disequilibrium and its use in mapping quantitative trait loci. The many causes of linkage disequilibrium can be understood as due to similarity in the coalescence tree of different loci. Consideration of the way this comes about allows us to divide linkage disequilibrium into 2 types: linkage disequilibrium between any 2 loci, even if they are unlinked, caused by variation in the relatedness of pairs of animals; and linkage disequilibrium due to the inheritance of chromosome segments that are identical by descent from a common ancestor. The extent of linkage disequilibrium due to the latter cause can be logically measured by the chromosome segment homozygosity which is the probability that chromosome segments taken at random from the population are identical by descent. This latter cause of linkage disequilibrium allows us to map quantitative trait loci to chromosome regions. The former cause of linkage disequilibrium can cause artefactual quantitative trait loci at any position in the genome. These artefacts can be avoided by fitting the relatedness of animals in the statistical model used to map quantitative trait loci. In the future it may be convenient to estimate this degree of relatedness between individuals from markers covering the whole genome. The statistical model for mapping quantitative trait loci also requires us to estimate the probability that 2 animals share quantitative trait loci alleles at a particular position because they have inherited a chromosome segment containing the quantitative trait loci identical by descent. Current methods to do this all involve approximations. Methods based on concepts of coalescence and chromosome segment homozygosity are useful, but improvements are needed for practical analysis of large datasets. Once these probabilities are estimated they can be used in flexible linear models that conveniently combine linkage and linkage disequilibrium information.