Application of single-step genomic evaluation for crossbred performance in pig

• Published in: Journal of animal science Vol. 94; no. 3; pp. 936 - 948
• Main Authors: Xiang, T, Nielsen, B, Su, G, Legarra, A, Christensen, O F
• Format: Journal Article
• Language: English
• Published: United States American Society of Animal Science 01.03.2016
• Subjects: Animals; Breeding; Computer Science; Female; Genome; Genomics - methods; Genotype; Hybridization, Genetic; Life Sciences; Litter Size - genetics; Male; Models, Genetic; Reproducibility of Results; Swine - genetics; Swine - physiology; single-step method; genetic correlation; crossbred performance; genomic evaluation; reliability; pig
• ISSN: 0021-8812; 1525-3163
• DOI: 10.2527/jas.2015-9930

Crossbreding is predominant and intensively used in commercial meat production systems, especially in poultry and swine. Genomic evaluation has been successfully applied for breeding within purebreds but also offers opportunities of selecting purebreds for crossbred performance by combining information from purebreds with information from crossbreds. However, it generally requires that all relevant animals are genotyped, which is costly and presently does not seem to be feasible in practice. Recently, a novel single-step BLUP method for genomic evaluation of both purebred and crossbred performance has been developed that can incorporate marker genotypes into a traditional animal model. This new method has not been validated in real data sets. In this study, we applied this single-step method to analyze data for the maternal trait of total number of piglets born in Danish Landrace, Yorkshire, and two-way crossbred pigs in different scenarios. The genetic correlation between purebred and crossbred performances was investigated first, and then the impact of (crossbred) genomic information on prediction reliability for crossbred performance was explored. The results confirm the existence of a moderate genetic correlation, and it was seen that the standard errors on the estimates were reduced when including genomic information. Models with marker information, especially crossbred genomic information, improved model-based reliabilities for crossbred performance of purebred boars and also improved the predictive ability for crossbred animals and, to some extent, reduced the bias of prediction. We conclude that the new single-step BLUP method is a good tool in the genetic evaluation for crossbred performance in purebred animals.