Large-scale single-step genomic evaluation for milk production traits

Abstract

The single-step method of genomic evaluation for milk volume, fat yield and protein yield was applied to the national dairy herd of New Zealand.  Genomic information from a 50K SNP chip was available on 5402 Holstein Friesian (HF), Jersey (J) and  HFxJ sires.  The genomic relationship matrix (GRM) or the Euclidean distance matrix (EDM) in a Gaussian kernel was used to augment the pedigree-based relationship matrix in the mixed model equations.  Scale parameters of 0.3, 0.5 and 0.7 were used for the GRM and 0.5, 0.7 and 0.9 for the EDM.   Traditional breeding values (BVs) were compared to genomic breeding values (GBVs) in two youngest cohorts of the progeny-tested sires (N=525).  An increasing scale parameter was associated with an increased inflation of the GBVs.  The EDM resulted in  lower inflation of fat GBVs than the GRM.  The effect was smaller and more variable for the other traits.   Augmenting the relationship matrix with the GRM versus the EDM and changing the magnitude of the scale parameters had little impact on the accuracy of the evaluation.