Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Animal Genomics and Improvement Laboratory » Research » Publications at this Location » Publication #381707

Research Project: Improving Dairy Animals by Increasing Accuracy of Genomic Prediction, Evaluating New Traits, and Redefining Selection Goals

Location: Animal Genomics and Improvement Laboratory

Title: Genomic heritability and prediction accuracy of additive and nonadditive effects for daughter pregnancy rate in crossbred dairy cows

item LIANG, ZUOXIANG - University Of Minnesota
item PRAKAPENKA, DZIANIS - University Of Minnesota
item Vanraden, Paul
item DA, YANG - University Of Minnesota

Submitted to: Journal of Dairy Science
Publication Type: Abstract Only
Publication Acceptance Date: 4/8/2021
Publication Date: 6/28/2021
Citation: Liang, Z., Prakapenka, D., Van Raden, P.M., Da, Y. 2021. Genomic heritability and prediction accuracy of additive and nonadditive effects for daughter pregnancy rate in crossbred dairy cows [abstract]. Journal of Dairy Science. 104(Suppl. 1):224(abstr. P168).

Interpretive Summary:

Technical Abstract: Genomic heritability and prediction accuracy of epistasis effects for daughter pregnancy rate (DPR) were evaluated using 79,294 SNPs and 9565 crossbred dairy cows. Heritability was estimated under the model with SNP additive effects (A), SNP dominance effects (D), and second and third order epistasis effects of A×A, A×D, D×D, A×A×A, A×A×D, A×D×D and D×D×D. Heritability estimate was 0.162 for additive effects, 0.283 for dominance effects, 0.555 for A×A effects, and zero or nearly zero for A×D, D×D and the third order epistasis effects. Genomic prediction included all SNP and epistasis effects with heritability greater than one percent, resulting in the prediction model with A, D and A×A effects only. Prediction accuracy as correlation between the genomic best linear unbiased prediction and the phenotypic values from a ten-fold validation study for each model was 0.268 for A-model, 0.363 for D-Model, 0.438 for A×A, 0.446 for A+D, 0.455 for A+(A×A), 0.467 for D+(A×A), and 0.475 for A+D+(A×A). Relative to the A-model, the D-model increased the prediction accuracy by 35.6%, A×A by 63.6%, A+D by 66.4%, A+(A×A) by 69.8%, D+(A×A) by 74.3%, and A+D+(A×A) by 77.2%. The heritability estimates and prediction accuracies showed that A×A effects were the largest contributor to DPR heterosis, followed by dominance and additive effects. The high additive heritability (0.16) in crossbred dairy cows relative to the low additive heritability in Holstein cows (0.025 according to our own estimate) indicated that a larger collection of favorable alleles from different breeds in crossbreds than in purebreds was part of the genetic mechanism underlying DPR heterosis. Combined with dominance and A×A effects, the results in this study support our GWAS finding in a separate study that genome-wide additive and nonadditive effects were the genetic mechanism of reproductive heterosis.