ARS | Publication request: Incorporating molecular breeding values with variable call rates into genetic evaluations

Submitted to: Journal of Animal Science Supplement
Publication Type: Abstract Only
Publication Acceptance Date: February 23, 2011
Publication Date: July 1, 2011
Citation: Kachman, S.D., Bennett, G.L., Hanford, K.J., Kuehn, L.A., Pollak, E.J., Snelling, W.M., Spangler, M.L., Thallman, R.M. 2011. Incorporating molecular breeding values with variable call rates into genetic evaluations [abstract]. Journal of Animal Science Supplement. 89(E-Supplement 1):164. (Abstract # 34).

Technical Abstract: A partial genotype for an animal can result from panels with low call rates used to calculate a molecular breeding value. A molecular breeding value can still be calculated using a partial genotype by replacing the missing marker covariates with their mean value. This approach is expected to change the distribution of the molecular breeding value by reducing the variance and percentage of genetic variation explained. Under independence, the variance of a molecular breeding value is twice the sum across all markers of the allele effects squared times the product of allele frequency and 1 minus allele frequency. For an animal, the proportion of the complete genotype molecular breeding value variance accounted for by a partial genotype is then the ratio of the variances calculated by summing over the partial and the complete set of markers. Similarly, the genetic covariance between a trait and a partial genotype molecular breeding value is also proportional to this ratio. If missing markers are selected at random, the proportion of complete genotype covariance between two animals with partial genotypes is proportional to the product of the ratios for each animal. A weaning weight molecular breeding value was constructed from 159 Single Nucleotide Polymorphisms using genotype and weaning weight data for 3,327 calves from the U.S. Meat Animal Research Center. Genotypes for 2,503 and weaning weights for 148,897 calves of seven pure breeds were used to evaluate the molecular breeding values. The genotype data had an average call rate of 85.2 percent. Two trait analyses were run for each breed separately, both with and without the effect of partial genotype variance ratios in the model. Overall, molecular breeding value heritability and its genetic correlation with weaning weight were larger when the effects of partial genotypes were incorporated into the analysis.