Abstract

A national genetic evaluation called the New Zealand Genetic Evaluation (NZGE) is undertaken weekly to support the genetic improvement of sheep via selective breeding in New Zealand (NZ). This is based on single step genomic best linear unbiased predictions (ssGBLUP), which allows animals that have genotypes and pedigree, or just pedigree, to be included in the same analysis (or evaluation). In theory, an optimal strategy is that the density and quality of SNPs used should be high enough to track all the haplotypes segregating in the population to be evaluated. Since genomic predictions have been used in NZ sheep, the content of the SNP panel used has decreased to a panel of 41K; despite that there are new SNP panels with more content available.. The effect of using this additional content and additional SNPs from an Ovine 600K chip was investigated using the trait fleece weight at 12-months. We found that a high density panel provided more accurate predictions. However, a SNP panel based on the 50K SNPs plus a relatively small subset of additional SNPs (e.g. from the 600K Chip), were as predictive as using the high density panel. We conclude that increased SNP density, beyond the 41K selected from the Ovine 50K SNP, chip will improve prediction accuracy. Our results suggest that the inclusion of SNPs from putative quantitative trait loci (QTL) may also improve prediction accuracy. However, further analyses will be helpful to distinguish the relative benefit to accuracy of increased SNP density across the genome and/or inclusion of SNPs associated with QTL. Keywords: sheep, genetics, genomics, breeding, ssGBLUP, genomic prediction

AM, Lee, KG Dodds, S-AN Newman, D Campbell, and SM Clarke

New Zealand Journal of Animal Science and Production, Volume 81, Online, 117-121, 2021