513
Accuracy of Molecular Breeding Values for Production and Efficiency Traits of Canadian Crossbred Beef Cattle Using a Cross-Validation Approach

Monday, August 18, 2014
Posters (The Westin Bayshore)
Everestus C. Akanno , University of Alberta, Edmonton, AB, Canada
Graham Plastow , University of Alberta, Edmonton, AB, Canada
Changxi Li , University of Alberta, Edmonton, AB, Canada
Stephen P Miller , Centre for Genetic Improvement of Livestock, University of Guelph, Guelph, ON, Canada
John A. Basarab , Alberta Agriculture and Rural Development, Lacombe, AB, Canada
Abstract Text:

A total of 4,218 animals genotyped for 50 K SNPs were used to evaluate accuracy of molecular breeding values (MBVs) of Canadian crossbred beef cattle traits. A cross-validation approach that involves random sampling of all animals to form 5-groups was applied. One group was dedicated as a validation set and the remaining four groups were combined to form a reference set. Genomic best linear unbiased prediction method was applied for training and prediction of MBVs in the validation set. Estimated breeding values (EBV) were used as the phenotype and a bivariate animal model was applied to each trait to estimate the genetic correlation between MBV and EBV. Accuracy of MBVs ranged from 0.05 to 0.68 for the studied traits. The results suggest that genomic predictions can be produced for crossbred beef cattle with greater efficiency for selection and management especially in younger animals.

Keywords:

feed efficiency

growth

SNP

See more of: Posters: Prediction using molecular information (Group 4)
See more of: Genetic Improvement Programs: Selection using molecular information
<< Previous Abstract | Next Abstract >>