Accuracy of genomic predictions for feed efficiency traits of beef cattle using 50K and imputed HD genotypes

Lu, D., Akanno, E. C., Crowley, J. J., Schenkel, F., Li, H., De Pauw, M., Moore, S. S., Wang, Z., Li, C., Stothard, P., Plastow, G., Miller, S. P. and Basarab, J. A. (2016) Accuracy of genomic predictions for feed efficiency traits of beef cattle using 50K and imputed HD genotypes. Journal of Animal Science, 94 4: 1342-1353. doi:10.2527/jas2015-0126


 
Related Publications and Datasets
 
Author Lu, D.
Akanno, E. C.
Crowley, J. J.
Schenkel, F.
Li, H.
De Pauw, M.
Moore, S. S.
Wang, Z.
Li, C.
Stothard, P.
Plastow, G.
Miller, S. P.
Basarab, J. A.
Title Accuracy of genomic predictions for feed efficiency traits of beef cattle using 50K and imputed HD genotypes
Journal name Journal of Animal Science   Check publisher's open access policy
ISSN 0021-8812
1525-3163
Publication date 2016-04-01
Sub-type Article (original research)
DOI 10.2527/jas2015-0126
Open Access Status Not yet assessed
Volume 94
Issue 4
Start page 1342
End page 1353
Total pages 12
Place of publication Champaign, IL, United States
Publisher American Society of Animal Science
Language eng
Formatted abstract
The accuracy of genomic predictions can be used to assess the utility of dense marker genotypes for genetic improvement of beef efficiency traits. This study was designed to test the impact of genomic distance between training and validation populations, training population size, statistical methods, and density of genetic markers on prediction accuracy for feed efficiency traits in multibreed and crossbred beef cattle. A total of 6,794 beef cattle data collated from various projects and research herds across Canada were used. Illumina BovineSNP50 (50K) and imputed Axiom Genome-Wide BOS 1 Array (HD) genotypes were available for all animals. The traits studied were DMI, ADG, and residual feed intake (RFI). Four validation groups of 150 animals each, including Angus (AN), Charolais (CH), Angus-Hereford crosses (ANHH), and a Charolais-based composite (TX) were created by considering the genomic distance between pairs of individuals in the validation groups. Each validation group had 7 corresponding training groups of increasing sizes (n = 1,000, 1,999, 2,999, 3,999, 4,999, 5,998, and 6,644), which also represent increasing average genomic distance between pairs of individuals in the training and validations groups. Prediction of genomic estimated breeding values (GEBV) was performed using genomic best linear unbiased prediction (GBLUP) and Bayesian method C (BayesC). The accuracy of genomic predictions was defined as the Pearson’s correlation between adjusted phenotype and GEBV (r), unless otherwise stated. Using 50K genotypes, the highest average r achieved in purebreds (AN, CH) was 0.41 for DMI, 0.34 for ADG, and 0.35 for RFI, whereas in crossbreds (ANHH, TX) it was 0.38 for DMI, 0.21 for ADG, and 0.25 for RFI. Similarly, when imputed HD genotypes were applied in purebreds (AN, CH), the highest average r was 0.14 for DMI, 0.15 for ADG, and 0.14 for RFI, whereas in crossbreds (ANHH, TX) it was 0.38 for DMI, 0.22 for ADG, and 0.24 for RFI. The r of GBLUP predictions were greatly reduced with increasing genomic average distance compared to those from BayesC predictions. The results indicate that 50K genotypes, used with BayesC, are more effective for predicting GEBV in purebred cattle. Imputed HD genotypes found utility when dealing with composites and crossbreds. Formulation of a fairly large training set for genomic predictions in beef cattle should consider the genomic distance between the training and target populations.
Q-Index Code C1
Q-Index Status Provisional Code
Institutional Status UQ

Document type: Journal Article
Sub-type: Article (original research)
Collections: HERDC Pre-Audit
Queensland Alliance for Agriculture and Food Innovation
 
Versions
Version Filter Type
Citation counts: TR Web of Science Citation Count  Cited 4 times in Thomson Reuters Web of Science Article | Citations
Scopus Citation Count Cited 2 times in Scopus Article | Citations
Google Scholar Search Google Scholar
Created: Sun, 05 Jun 2016, 10:21:37 EST by System User on behalf of Learning and Research Services (UQ Library)