Genomic predictions for yield traits in US Holsteins with unknown parent groups

The objective of this study was to assess the reliability and bias estimated breeding values (EBV) from traditional BLUP with unknown parent groups (UPG), genomic EBV (GEBV) single-step (ssGBLUP) UPG for pedigree relationship matrix (A) only (SS_UPG), GEBV ssGBLUP both A among genotyped animals (A22; SS_UPG2) using 6 large phenotype-pedigree truncated Holstein data sets. complete included 80 million records milk, fat, protein yields 31 cows recorded since 1980. Phenotype-pedigree truncation scenarios phenotypes before 1990 2000 combined information after 2 or 3 ancestral generations. total 861,525 bulls progeny phenotypic were used in analyses. Reliability (inflation/deflation) obtained 2,710 based on deregressed proofs, 381,779 born 2014 predictivity (adjusted cow phenotypes). reliabilities young varied 0.29 0.30 across traits unaffected by number generations pedigree. Reliabilities ranged 0.54 0.69 SS_UPG slightly affected truncation. 0.73 SS_UPG2 regression coefficient bull proofs (G)EBV (i.e., EBV) 0.86 0.90 BLUP, 0.77 0.94 SS_UPG, 1.00 ± 0.03 SS_UPG2. Cow 0.22 0.28 0.48 0.51 highest predictivities most extreme truncation, whereas SS_UPG2, also truncations. 1.02 0.02 which indicated least biased predictions. Computations set took 17 h 58 23 same computations 7, 36, 15 h, respectively. converged fewer rounds than up twice as many rounds. Thus, assigned A22 provided accurate unbiased evaluations, regardless scenario. Old (before set) did not affect predictions selection candidates, especially