|CESARANI, ALBERTO - University Of Georgia|
|MASUDA, YUTAKA - University Of Georgia|
|TSURUTA, SHOGO - University Of Georgia|
|NICOLAZZI, EZEQUIEL - Council On Dairy Cattle Breeding|
|LOURENCO, DANIELA - University Of Georgia|
|MISZTAL, IGNACY - University Of Georgia|
Submitted to: Journal of Dairy Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/18/2020
Publication Date: 5/1/2021
Citation: Cesarani, A., Masuda, Y., Tsuruta, S., Nicolazzi, E.L., Van Raden, P.M., Lourenco, D., Misztal, I. 2021. Genomic predictions for yield traits in US Holsteins with unknown parent groups. Journal of Dairy Science. 104(5):5843–5853. https://doi.org/10.3168/jds.2020-19789.
Interpretive Summary: Reliable and unbiased genomic predictions depend on various factors including compatibility between pedigree and genomic information. Moderate levels of pedigree missingness can compromise genomic predictions if unknown pedigree data are not correctly modeled. We investigated single-step genomic BLUP models that fitted unknown parent groups either for the pedigree relationship matrix only (SS_UPG) or for both the pedigree relationship matrix and the relationship matrix among genotyped animals (SS_UPG2). The SS_UPG2 genomic predictions for yield traits in US Holstein were highly reliable and nearly unbiased for young bulls and cows. Removal of old phenotypes and pedigree had no impact on SS_UPG2 genomic predictions.
Technical Abstract: The objective of this study was to assess the reliability and bias of EBV from traditional BLUP with unknown parent groups (UPG), GEBV from ssGBLUP with UPG for the pedigree relationship matrix (A) only (SS_UPG), and GEVB from ssGBLUP with UPG for both A and the relationship matrix among genotyped animals (A_22) (SS_UPG2) using six large phenotype-pedigree truncated Holstein datasets. The complete data included 80 million records for milk, fat, and protein yield from 31 million cows born since 1980. Phenotype-pedigree truncation scenarios included truncation of phenotypes for cows born before 1990 and 2000 combined with truncation of pedigree information after 2 or 3 ancestral generations. A total of 861,525 genotyped bulls with progeny and cows with phenotypic records were used in the analyses. Reliability and bias (inflation/deflation) of GEBV were obtained for 2,710 bulls based on deregressed proofs, and on 381,779 cows born after 2014 based on predictivity (adjusted cow phenotypes). BLUP reliabilities for young bulls varied from 0.29 to 0.30 across traits and were unaffected by data truncation and number of generations in the pedigree. Reliabilities ranged from 0.54 to 0.69 for SS_UPG and were slightly affected by phenotype-pedigree truncation. Reliabilities ranged from 0.69 to 0.73 for SS_UPG2 and were unaffected by phenotype-pedigree truncation. The regression coefficient of bull deregressed proofs on (G)EBV ranged from 0.86 to 0.90 for BLUP, 0.77 to 0.94 for SS_UPG, and was 1.00 +- 0.03 for SS_UPG2. Cow predictivity ranged from 0.22 to 0.28 for BLUP, 0.48 to 0.51 for SS_UPG, and 0.51 to 0.54 for SS_UPG2. The highest cow predictivities for BLUP were obtained with the most extreme truncation, whereas SS_UPG2 cow predictivities were unaffected by phenotype-pedigree truncations. The regression coefficient of cow predictivities on (G)EBV was 1.02 +- 0.02 for SS_UPG2 with the most extreme truncation, which indicated the least biased predictions. Computations with the complete dataset took 17h with BLUP, 58h with SS_UPG, and 23h with SS_UPG2. The same computations with the most extreme phenotype-pedigree truncation took 7h, 36h, and 15h. The SS_UPG2 converged in fewer rounds than BLUP, whereas SS_UPG took up to twice as many rounds. Thus, the ssGBLUP with unknown parent groups assigned to both A and A_22 provided accurate and unbiased evaluations regardless of phenotype-pedigree truncation scenario. Old phenotypes (before 2000 in this dataset) did not impact the reliability of predictions for young selection candidates, especially in SS_UPG2.