Cross-validation of single-step genetic evaluation in U.S. Katahdin sheep

Abstract

<p>Genomic information is used in genetic evaluation to improve prediction accuracy in most livestock species. Such is less so in sheep, particularly in the U.S. In this study, the impact of implementing singlestep genomic BLUP as compared to pedigree BLUP for weight and faecal egg count traits was evaluated in U.S. Katahdin sheep. Two methods of cross validation were utilised to compare the predictive ability and bias of estimated breeding values of these methods. Genomic information improved predictive ability for both traits, and reduced bias in the evaluation of faecal egg counts. Accuracies of estimated breeding values improved appreciably in genotyped animals. The benefit from including genomic information based on cross validation was minimal but is expected to improve as the reference population grows. Single-step genomic BLUP procedures developed will be used to update those applied in the routine genetic evaluation offered through the U.S. National Sheep Improvement Program.</p>