Impact of using a genomic model (ssGBLUP) vs. a traditional model (PBLUP) on the accuracy of breeding values for parasite resistance in Australian Merino in Uruguay

Abstract

Resistance to gastrointestinal parasites is a priority trait in sheep breeding programs due to its impact on animal health. The objective of this study was to compare the accuracy of genetic evaluations using a pedigree-based model (PBLUP) and a genomic-based model (ssGBLUP) in Australian Merino sheep for the trait eggs per gram of fecal matter (EPM), an indirect measure of parasite burden. A multivariate model with EPM1, EPM2, and shearing weight as variables was used. This study presents EPM1 results, with phenotypic and genealogical data from 107,250 animals; 37,575 with EPM1 records and 3,981 genotyped and imputed to the Axiom® Ovine 60K panel. A total of 3,633 animals born between 2001 and 2023 had EPM records and were genotyped. The accuracies of the breeding values obtained with both models were compared. The results are presented separately into four groups of animals according to the availability of genotypes and phenotypes: genotyped with data (G-C/D), genotyped without data (G-S/D), not genotyped with data (NG-C/D), and not genotyped without data (NG-S/D). The ssGBLUP model showed higher accuracies for all the groups analyzed. The greatest increase was obtained for G-S/D (14.8%), with accuracies ranging from 0.61 to 0.70. The lowest impact was in NG-C/D (1.5%), from 0.65 to 0.66. In G-C/D it was 0.72 to 0.76 (5.6%), and in NG-S/D, from 0.26 to 0.28 (7.7%). These results show that the use of genomic information improves the accuracy of breeding values for HPG, especially in animals without phenotypic records. It is concluded that the ssGBLUP model offers advantages for the genetic evaluation of parasite resistance in the Australian Merino breed.