Single step genomic evaluations for the Nordic Red Dairy cattle test day data

Abstract

Most genomic evaluations are currently based on multi step -approach that requires 1) traditional evaluation with an animal model, 2) extraction of pseudo-observations, and 3) the genomic model to predict direct genomic values (DGV) of candidate animals without own records. In the single step analysis the phenotypic records are combined directly with genomic information, and the resulting genomic enhanced breeding value (GEBV) already combine both sources of information optimally. The objectives of this study were to evaluate the feasibility of the TD single step model using phenotypic records of Nordic RDC cows, and to quantify the accuracy when using single step TD model. The results show that the use of phenotypic test-day records in single step analysis is realistic and easy to implement. Moreover, single step TD models give comparable results to original TD models and considerably higher GEBV validation reliabilities and validation regression coefficients. This indicates that inflation is smaller than with DGVs from sire model validations although it still exists. Thus, it provides a good alternative to the current multi step -approach.