PAG-XVII  Plant & Animal Genomes XVII Conference

January 10-14, 2009
Town & Country Convention Center
San Diego, CA



W080 : Cattle/Sheep


Prospects For Whole Genome Selection In Beef Cattle

Jeremy F Taylor , Stephanie D McKay , Mathew C McClure , Jae W Kim , Dan A Vasco , Jared E Decker , Megan M Rolf , Robert D Schnabel

  Division of Animal Sciences University of Missouri Columbia, MO 65211

A population of 1,958 registered Angus bulls has been genotyped using the Illumina BovineSNP50 assay. A total of 41,028 SNP passed QA/QC and possessed MAF>5%. Whole genome association analyses have been conducted for the 17 traits for which EPD are computed from phenotypic data and genome-wide significance levels were established by permutation analysis. At genome-wide P<0.001 we found as few as 182 associations with Fat Thickness to 12,066 associations with Yearling Weight. Since multiple SNP are associated with each QTL we are developing approaches to reduce the number of SNP required to ‘tag’ each QTL and allow the prediction of genetic merit. For example, the 509 SNP associated with Marbling can be filtered to 129 QTL which explain 59.1% of the variation in Marbling EPD. The importance of this approach stems from the fact that the BovineSNP50 assay is unlikely to be widely used for genome wide selection in beef cattle due to cost. Assays with many fewer SNP that can be commercialized at low cost will be required for deep technology penetration within the beef industry. Despite this, it appears likely that several thousands of SNP must be tested to explain 50-60% of the variation in all 17 traits. Additional populations including 1,080 registered Limousin have also been genotyped to allow the development of across breed tests. It is not yet clear if the BovineSNP50 marker density will be sufficient to develop tests with utility across all European and British Bos taurus breeds.