Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Animal Genomics and Improvement Laboratory » Research » Publications at this Location » Publication #338519

Title: Genomic predictions for crossbreds from all-breed data

item Tooker, Melvin
item Vanraden, Paul
item Fok, Gary

Submitted to: Journal of Dairy Science
Publication Type: Abstract Only
Publication Acceptance Date: 2/27/2017
Publication Date: 6/24/2017
Citation: Tooker, M.E., Van Raden, P.M., Fok, G.C. 2017. Genomic predictions for crossbreds from all-breed data. Journal of Dairy Science. 100(Suppl. 2):409–410(abstr. 461).

Interpretive Summary:

Technical Abstract: Genomic predictions of transmitting ability (GPTAs) for crossbred animals were computed from marker effects of 5 dairy breeds weighted by each breed’s genomic contribution to the crossbreds. Estimates of genomic breed composition are labeled breed base representation (BBR) and are reported since May 2016 for all 1.6 million genotyped dairy animals. Animals with >94% of any breed were rounded to 100%, and contributions of other breeds were set to 0%. All-breed scale GPTAs were first computed for each pure breed for traits milk, fat, protein, productive life, somatic cell score, daughter pregnancy rate, cow conception rate, livability, and net merit. These estimates included foreign information from multi-trait across-country evaluation (MACE) and foreign dams converted from within-breed to the all-breed base. Then, marker effects for each breed were blended by BBR to compute evaluations for crossbreds (< 94% purebred) for those same traits. Conformation traits do not have an all-breed scale, so only the Jersey marker effects were applied to the crossbreds, and results seemed reasonable. Calving traits are not predicted for crossbreds, and instead a common mean was used for all crossbreds as is the current practice for breeds other than Holstein and Brown Swiss. All-breed GPTAs were then converted to within-breed GPTAs. Correlations of GPTAs for purebreds computed on the all-breed vs. current within-breed scales were 0.97 to 0.99 for most traits and breeds. Crossbred GPTAs were then computed for 44,023 crossbreds, 20,367 of which had no previous GPTAs because of breed check edits. The new GPTAs were for 1,822 Jersey x Holstein crossbreds with >40% of both breeds (F1 crosses), 75 Brown Swiss x Holstein F1, 7,237 Holstein backcrosses with >67% and <94% Holstein, 7,820 Jersey backcrosses, 313 Brown Swiss backcrosses, 1,763 other crossbreds of various mixtures, and 1,337 purebreds that had previously failed breed checks. Additional automation and redesign of many downstream programs is required for the new all-breed system to be used in weekly, monthly, and full releases. The new system is expected to provide accurate predictions for crosses among the 5 dairy breeds evaluated.