Short communication: Relationship of call rate and accuracy of single nucleotide polymorphism genotypes in dairy cattle

Authors: Cooper, Tabatha; Wiggans, George; Vanraden, Paul

Submitted to: Journal of Dairy Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/17/2013
Publication Date: 5/1/2013
Citation: Cooper, T.A., Wiggans, G.R., Van Raden, P.M. 2013. Short communication: Relationship of call rate and accuracy of single nucleotide polymorphism genotypes in dairy cattle. Journal of Dairy Science. 96(5):3336-3339.

Interpretive Summary: Genotype accuracy is important for genomic evaluation. Call rate effects SNP genotype accuracy and ability to validate parentage. Although opposite homozygote errors are usually not observed in genotypes until call rate is below 80%, for genotypes below 90% the common error was a single allele change. To detect that a heterozygous allele is wrong, both parents must be genotyped. Errors were determined by comparison of a low call rate genotype with a high call rate genotype for the same animal. For the genotype to have accuracy above 99% the call rate must be above 90%.

Technical Abstract: Call rate has been used as a measure of quality on both a single nucleotide polymorphism (SNP) and animal basis since SNP genotypes were first used in genomic evaluation of dairy cattle. The genotyping laboratories perform initial quality control screening and genotypes that fail are usually excluded from routine delivery of data. However, laboratories may submit lower call rate genotypes for research, or to be passed on for use in other country’s evaluations with differing edit limits. The collection of these genotypes has allowed for the study of how call rate affects accuracy of genotype calls. The average call rate for usable Holstein, Jersey and Brown Swiss animals is 99.6% with less than 1% of genotypes submitted for evaluation rejected due to low call rate. Call rate effects SNP genotype accuracy and ability to validate parentage. Although an opposite homozygote error is usually not observed in genotypes until call rate is below 80%, for genotypes below 90% the common error was a single allele change e.g. the homozygous allele to the heterozygous allele or the reverse. When both parents have been genotyped some heterozygous allele errors can be detected. In this study, all SNP could be checked because a low call rate genotype was compared with a high call rate one for the same animal. For the genotype to have accuracy above 99% the call rate must be above 90%.