ARS | Publication request: Development and Applications of a Bovine 50,000 SNP Chip

Submitted to: International Symposium of Animal Functional Genomics
Publication Type: Abstract
Publication Acceptance Date: March 15, 2008
Publication Date: April 6, 2008
Citation: Taylor, J.F., Sonstegard, T.S., Van Tassell, C.P., Matukumalli, L., Smith, T.P., Allen, M.L. 2008. Development and Applications of a Bovine 50,000 SNP Chip. International Symposium of Animal Functional Genomics.

Technical Abstract: To develop an Illumina iSelect high density single nucleotide polymorphism (SNP) assay for cattle, the collaborative iBMC (Illumina, USDA ARS Beltsville, University of Missouri, USDA ARS Clay Center) Consortium first performed a de novo SNP discovery project in which genomic reduced representation libraries were deep sequenced using a next generation short read technology. This study produced 62,042 putative SNP which simultaneously had minor allele frequencies (MAF) estimated and were mapped to Btau3.1 coordinates. Genotype data for 66 animals used in the SNP discovery project validated 91% of the predicted SNP and revealed a correlation of 67% between sequence- and genotype-based estimates of allele frequency. Subsequently, all publicly available SNP were considered for inclusion on the assay based upon their Btau3.1 spacing, MAF, designability scores and probable conversion rates. The 60,800 available assay probes were allocated to 58,336 SNP of which 23,521were derived from the discovery project. A total of 53,349 loci were found to be high quality scoreable SNP and 43,578 have MAF=0.05 in the U.S. Angus and Holstein populations. The INRA 3,000-rad and Texas A&M BovR12 12,000-rad radiation hybrid (RH) panels and approximately 8,000 halfsibs have been genotyped to produce RH, linkage, and integrated maps which we expect will assist correct errors in the Btau4.1 assembly and successfully order the majority of scoreable SNPs on the assay. The assay is being used by the iBMC members to test the feasibility of Genome-Wide Selection (GWS) strategies in U.S. Angus, Limousin, and Holstein populations and for the joint linkage and linkage disequilibrium mapping of Quantitative Trait Loci in a variety of beef and dairy populations with phenotypes that include those for which EPD/PTA are routinely calculated and also for feed efficiency, meat tenderness, and palatability. Preliminary results of these analyses will be presented at the meeting. In addition, we have used high density SNP data to estimate individual multilocus average homozygosity and pairwise average allele sharing among 23 registered Angus cattle. Additive coefficients of relationship and inbreeding coefficients were estimated for these animals using a 62 generation pedigree comprising 20,974 registered animals. We found linear relationships between individual homozygosity and pedigree inbreeding coefficients (r=0.80) and between pairwise allele sharing and additive coefficients of relationship (r=0.93). These results suggest that molecular data may be used to accurately reconstruct the numerator relationship matrix among population members which has profound consequences for the management of endangered species and the prediction of additive genetic merit using mixed linear model procedures. It appears clear that the applications of high density SNP genotyping to animal management and breeding programs are numerous and that these are likely to be of profound impact to animal agriculture. We anticipate rapid adoption of these technologies within the international beef and dairy cattle industries.