Multibreed genomic evaluations using purebred Holsteins, Jerseys, and Brown Swiss

Area of Science:

• Animal Genetics and Breeding
• Quantitative Genetics
• Dairy Cattle Improvement

Background:

• Traditional US Department of Agriculture (USDA) dairy cattle genetic evaluations utilize multibreed models for yield and health traits.
• Current genomic evaluations predominantly rely on within-breed models, despite the established use of multibreed approaches in traditional methods.
• The study addresses the need to integrate multibreed strategies into genomic evaluations for enhanced accuracy.

Purpose of the Study:

• To develop and evaluate multibreed genomic models for dairy cattle genetic evaluations.
• To compare the performance of different multibreed model strategies against existing within-breed genomic models.
• To assess the accuracy of across-breed and within-breed genomic predicted transmitting ability (GPTA).

Main Methods:

• Developed and tested three multibreed genomic evaluation methods using 19,686 genotyped Holstein, Jersey, and Brown Swiss cattle.
• Data were divided into training (n=7,307) and validation (n=3,005) sets, utilizing 43,385 single nucleotide polymorphisms (SNP).
• Method 1: Breed-specific SNP effects applied across breeds. Method 2: Common SNP effects from combined data. Method 3: Correlated SNP effects using a multitrait model.

Main Results:

• Method 1 demonstrated poor performance, with low correlations (<30%) between direct genomic values from different breeds.
• Method 2 (across-breed GPTA) showed higher R(2) than parent average but generally lower than within-breed GPTA, except for some Brown Swiss traits.
• Method 3 yielded the highest adjusted R(2) for protein yield across all breeds, outperforming within-breed predictions, though improvements were modest (0.01-0.01).

Conclusions:

• Multibreed genomic evaluation, particularly Method 3, offers significant improvements over current within-breed genomic predictions for dairy cattle.
• The multitrait approach (Method 3) effectively leverages information across breeds, especially beneficial for smaller populations like the Brown Swiss.
• Further research and refinement of multibreed models are warranted to maximize their potential in dairy cattle genetic improvement programs.