Submitted to: Journal of Dairy Science
Publication Type: Abstract Only
Publication Acceptance Date: 2/24/2011
Publication Date: 6/30/2011
Citation: Wiggans, G.R., Cooper, T.A., Olson, K.M., Van Raden, P.M. 2011. Use of the Illumina Bovine3K BEAD chip in dairy genomic evaluation. Journal of Animal Science 89(E-Suppl. 1)/Journal of Dairy Science 94(E-Suppl. 1):420(abstr. 329). Interpretive Summary:
Technical Abstract: Genomic evaluations using genotypes from the Illumina Bovine3K BEAD chip became available in September 2010 and were made official in December 2010. Approximately 4,000 genotypes a month, (93% female) are from this low cost chip. To integrate the 3K genotypes into the evaluations, they are imputed from the 2,614 single nucleotide polymorphisms (SNP) used from the 3K chip to the 42,503 used in evaluation. Imputation is not perfect, so a reliability penalty is applied. The average 3K genomic evaluation reliability is 5 points lower than for 50K evaluations. The accuracy of imputation is largely dependent on an animal’s relationship to the genotyped population. The average imputed call rate for 3K genotypes is 95.2% and ranges from 71.0% to 97.0%. The animals that have a low imputed call rate are those who have unknown pedigree or no genotyped relatives. 94.5% of animals tested using the 50K chip have at least one genotyped parent whereas only 84.2% of 3K genotyped animals do. For approximately 8% of 3K genotypes the sire is determined to be incorrect. If the true sire of an animal is genotyped, it can be identified with >99% certainty. Other errors such as dam conflicts, unidentified identical twins / split embryos and breed conflicts also prevent genotypes from being used. The chemistry used for the 3K chip is different from that of the 50K chip and can cause greater variability in the accuracy of the genotypes. Because of this, a 3K specific check was added which excludes approximately 1% of samples. They are rejected because they have a high proportion of conflicts between SNP of the sire/dam and progeny when the parent is correct. The performance of SNP also differed between chips, resulting in 272 SNP that were usable on the 50K chip being not usable on 3K. The 3K chip has been successful in extending genotyping to a larger portion of the cow population. The evaluation system has been modified to accommodate the characteristics of the chip. With experience, the laboratories have improved the quality of their genotypes. Improvements in accuracy of imputation and other improvements will further improve the accuracy of 3K based genomic evaluations.