Farm Table says:
Accuracies from purebred and crossbred predictions in Australian beef cattle
What is the problem?
Genomic selection has been used by beef cattle breeders to select for desirable traits in cattle in order to improve overall herd performance. However, implementing genomic selection in Australian beef cattle is challenging due to the range of breeds and small numbers of genotyped individuals per breed. This paper displays accuracies of genomically estimated breeding values (GEBV) calculated from predictive equations from a training set of both pure-breed and cross-breed animals.
What did the research involve?
The Australian Beef Cooperative Research Centre (Beef CRC) developed genomic predictive equations from a set of animals that consisted of a range of breeds and crosses including Angus, Limousin and Brahman. GEBVs were calculated from these predictive equations, and their accuracies were determined by their genetic correlation to their phenotypic target trait. Traits targeted include post-weaning live weight, scan eye muscle area, scan rib fat, live weight at feedlot start and end, carcass rib fat, carcass intramuscular fat and carcass weight.
What were the key findings?
The accuracies for the majority of GEBVs for Angus and Brahman ranged from 0.1 to 0.4, with accuracies for abattoir traits generally higher than live animal traits. Whilst these accuracies are generally low, they correlate with other GEBVs from across-breed populations. No reasonable results could be achieved for the Limousin breed, for any trait. Overall, GEBV accuracies were highest when predictive equations from the cross-breed training population were used.
For beef cattle breeders wishing to implement genomic selection, the authors recommend using prediction equations derived from the Beef CRC cross-breed training population, but only if their breeds were part of that training population i.e. Angus, Limousin and Brahman.