Use of DNA fingerprints for the detection of major genes for quantitative traits in domestic species

Publisher Website
Google Scholar
Abstract
Summary. The detection of marker loci linked to major genes or quantitative trait loci (QTL) of large effect in farm animal populations is of great potential value, both because it allows the easy manipulation of the major genes and because it provides a possible route to their ultimate isolation. At present the number of markers available is limited in farm animals. DNA fingerprints provide a promising source of informative marker loci and have the advantage that several loci can be detected on a single Southern hybridization. The disadvantage of DNA fingerprints is the difficulty in determining allelism of DNA fingerprint bands in different pedigrees and the fact that not all potentially resolvable loci can be resolved in a single pedigree. With probes capable of detecting 50 randomly distributed loci, about 50% of the genome of a typical domestic mammal might be expected to be closely linked to a marker (at a distance of 0.2 Morgans or less). If a proportion of DNA fingerprint loci prove to be clustered near chromosomal telomeres or elsewhere in the genome, coverage will be less. In order to detect linkage to a major gene, sires known or suspected to be heterozygous are used to produce large half‐sibships, all animals in the pedigree are DNA fingerprinted and the phenotypes of the offspring are recorded. Where several heterozygous sires are available, sires can be selected in an attempt to maximize the number of marker loci resolved. The optimum number of sires needed to produce pedigrees will depend upon the size of the major gene, the number of DNA fingerprint probes available and the characteristics of the DNA fingerprints produced, but often one or two pedigrees will be optimum. Monte Carlo simulation was used to explore the power of detection of linkage between a major gene and a marker locus in a backcross. Maximum likelihood and analysis of variance of mean differences between marker genotypes were of similar power, but maximum likelihood provided reasonable estimates of the major gene effect and its linkage to the marker under some circumstances. One hundred offspring informative for the segregation of a marker would provide reasonable power for the detection of a gene causing a difference between the heterozygote and the homozygote of at least one within‐sire, within‐genotype standard deviation when linkage was very close (0.05 or less). With looser linkage (0.2 Morgans or less), a sample of the same size has reasonable power to detect genes causing at least two within‐sire, within‐genotype standard deviations difference between the heterozygote and the homozygote. The Booroola gene causes a difference of approximately 2.75 within‐sire, within‐genotype standard deviations for ovulation rate between heterozygous carriers and homozygous non‐carriers of the Booroola allele. DNA fingerprints can provide a worthwhile probability of finding a marker linked to the Booroola gene in a sample of less than 100 informative offspring.