Extensive Genome-wide Linkage Disequilibrium in Cattle

Abstract

A genome-wide linkage disequilibrium (LD) map was generated using microsatellite genotypes (284 autosomal microsatellite loci) of 581 gametes sampled from the dutch black-and-white dairy cattle population. LD was measured between all marker pairs, both syntenic and nonsyntenic. Analysis of syntenic pairs revealed surprisingly high levels of LD that, although more pronounced for closely linked marker pairs, extended over several tens of centimorgan. In addition, significant gametic associations were also shown to be very common between nonsyntenic loci. Simulations using the known genealogies of the studied sample indicate that random drift alone is likely to account for most of the observed disequilibrium. No clear evidence was obtained for a direct effect of selection ("Bulmer effect"). The observation of long range disequilibrium between syntenic loci using low-density marker maps indicates that LD mapping has the potential to be very effective in livestock populations. The frequent occurrence of gametic associations between nonsyntenic loci, however, encourages the combined use of linkage and linkage disequilibrium methods to avoid false positive results when mapping genes in livestock.Recently, linkage disequilibrium (LD) has received considerable attention as it may be exploited to more effectively map genes underlying both simple and complex (dichotomous and continuously distributed) traits (Terwilliger and Weiss 1998). The potential advantage of LD mapping over conventional linkage analysis performed within families lies in the use of "historical" recombinants, thereby increasing mapping resolution (e.g., Hästbacka et al. 1992;Talbot et al. 1999) and power. To be effective, however, LD-mapping requires a marker density compatible with the distances across which LD extends in the population of interest. Kruglyak (1999) estimated by simulation that useful levels of LD were unlikely to extend beyond an average distance of 3 kb in the human, thereby implying the need for a marker map comprising ∼500,000 SNPs Although experimental LD data are accumulating in the human (e.g., Laan and Pääbo 1997; Nickerson et al. 1998) and some primate species (Crouau-Roy et al. 1996), little is known about the extent of LD in most other mammals, including domestic species. In this paper, we have used genotypes obtained with a panel of 284 microsatellites to measure genome-wide LD in the dutch black-andwhite dairy cattle population. We make the remarkable observation that intrachromosomal LD extends over several tens of centimorgans, and that gametic phase disequilibrium is common between non syntenic loci.

Linkage disequilibrium on chromosome 6 in Australian Holstein-Friesian cattle

“…II). The extent of LD observed in this study in Australian dairy cattle is comparable to that reported for a Dutch Holstein population [10] in which LD between syntenic markers extended over "several tens of centimorgans". However, there were differences in the sizes of the sample and the marker density used in these two studies.…”

“…In the human genome, the extent of LD is much smaller (5-100 kb) than reported in this study and in the Dutch Holstein population [10]. Recently Hinds et al [16] reported the analysis of 1.58 million human SNP and suggested that the same power in association mapping of complex traits can be achieved by using around 300 000 SNP.…”

“…This statistic has been used to measure the extent of LD in human [3] and livestock populations in a range of studies (e.g. [10,28,34,42]), thereby facilitating the comparison of the results in this study with other studies. D allows LD to be measured with highly polymorphic markers and is less sensitive to variation in marker allele frequencies than other measures of LD [14,46].…”

“…The extent of LD in livestock is expected to be higher than in humans, since the forces producing LD (admixture, selection and small effective population sizes) are more extreme in livestock populations [33]. The evidence supporting this expectation is accumulating: high levels of LD have been found to extend for many centimorgans in dairy cattle [10], sheep [28], pigs [34] and horses [43]. These studies used family information to infer the most likely phase of the dams' haplotypes.…”

Mapping multiple QTL using linkage disequilibrium and linkage analysis information and multitrait data

“…Linkage disequilibria can be found over large distances [7], and thus pure LD analysis can easily result in the detection of false positives. It is therefore important to detected LD that are due to close linkage, and the combined use of linkage disequilibrium and linkage analysis information will avoid spurious long distance associations leading to false likelihood peaks, because the linkage analysis information will not confirm such spurious associations [19,21,22].…”