Inclusion of MACE proofs in single-step genomic analysis

Abstract

The integration of MACE proofs in single-step genomic analysis is important to provide the dairy industry with the best EBVs, especially in countries that import a major part of their genetics. The method developed earlier that uses Deregressed Proofs (DRP) that account for correlations between traits, but not relationships among MACE bulls was largely successful, but as we show here leads to large overestimations if MACE bulls are related. We developed an alternative approach that uses DRPs which take into account relationships among bulls, but still uses the old weights based on the assumption of no relationship. This proofed to be a better predictor of performance in Australia for bulls that had both a genotype and MACE proof partially based on their Australian daughters. An additional adjustment to account for that daughter information sent to Interbull proved ineffective, with the regression coefficient 0.82 in both cases. Bulls that were not expected to be affected by the singe-step procedure as they had no Australian daughters and no genotype, did in fact show large changes (regression coefficient 066), showing that the weights need to be in-line with the DRP estimation procedure.