Changes to evaluation system (August 2014)
Genomic inbreeding for mating programs
By Chuanyu Sun and Paul VanRaden
August 15 update: Files are provided on Friday instead of Thursday and contain genomic relationships, which must be divided by 2 to obtain genomic inbreeding.
A calf's inbreeding can be predicted more accurately using its parents' genotypes than their pedigrees. Files containing the genomic relationship of each genotyped female with each marketed male are now provided to industry cooperators for use in genomic mating programs. Predicted inbreeding of a calf equals one-half its parents' genomic relationship. A full file will be provided on the Thursdays after August, December, and April full releases, and update files will be provided at monthly releases containing only the new females not previously genotyped (as rows) with the previous list of marketed bulls (as columns). The marketed bulls most likely to be considered for mating will include those with a status of Active, Genomic, Foreign available, or Limited within the past 8 months, plus the highest of the newly Collected bulls, and Inactive bulls in the top 500 for net merit with a minimum production and type reliability of at least 95% (or 90% for Jerseys). The list of males provided in the monthly evaluations will match the previous full release so that concatenating the files of females will be simpler for use in mating programs. Separate files are provided for each breed; current size of the Holstein file is 5 Gbytes before and 1 Gbyte after being zipped. The format is blank separated and includes a first row containing the 17-byte IDs of the bulls, with each following line containing the 17-byte ID of the female and the predicted inbreeding coefficients when mated to each of the bulls. Mating programs that use genomic instead of pedigree inbreeding can improve economic merit by $30 per heifer calf and nearly pay for the cost of genotyping the dams. For herds currently using random mating, switching to an optimum genomic mating program will reduce genomic inbreeding by >3 percentage points and increase calf merit by $72 for Holsteins, $103 for Jerseys, and $67 for Brown Swiss. These benefits may greatly exceed the small increase in technician labor to find the correct straw for each cow at the time of insemination. This Journal of Dairy Science article provides further information.
Jersey haplotype 2 (JH2)
By Paul VanRaden, Dan Null, Jana Hutchison, Derek Bickhart, and Steve Schroeder
A second haplotype affecting fertility in Jerseys (JH2) was discovered from CDCB data and a third defect called Fertility1 was discovered in New Zealand (NZL) Jerseys by Livestock Improvement Corporation (LIC). Both JH2 and Fertility1 are located on chromosome 26 but at separate locations; the Fertility1 defect is at ~25 Mbase and appears only in NZL families, whereas JH2 is at 8.8-9.4 Mbase and has a long history in USA JE sires. The oldest genotyped JH2 carrier is JEUSA564125 Liberators Basilius born in 1954, but few current carriers inherited JH2 from him. Instead, most inheritance traces to the paternal half sibs JEUSA593883 S.S. Quicksilver of Fallneva born in 1960 and JEUSA596832 Observer Chocolate Soldier born in 1962, and also to JEUSA602283 Favorite Secret Triumph born in 1964. The carrier frequency was 14-28% in the decades before 1990 (similar to JH1), but has decreased steadily since 1990 to only 3.2% across all genotypes and only 2.6% now, much lower than the current JH1 frequency of 24%. The reason for the frequency decline is not clear, but is why JH2 took longer than JH1 to detect. Field JH2_PC reports pedigree confirmation for the JH2 haplotype using the same codes as JH1_PC: “C” for confirmed and “N” for not confirmed.
The estimated effect on conception rate of -4.0% ± 1.5% based on 1,164 JH2 carrier sire by JH2 carrier maternal grandsire matings is slightly larger than for JH1, but the average loss from random mating is much less because fewer mates are JH2 carriers. The embryos are lost by 60 days after insemination, similar to JH1, because estimated fertility losses after 60 days did not increase. The effect on stillbirth is very small and not significant based on 740 carrier sire by carrier MGS matings. The number of expected homozygous JH2 genotypes was 14, and none were found. Sequence data was examined for 3 JH2 carrier bulls that happened to have already been sequenced for JH1, but the coverage in the section around JH2 was poor and no obvious mutation was found, so further sequencing is needed to find the mutation.
A haplotype test for Fertility1 cannot yet be developed because the CDCB genotype database contains very few Fertility1 carriers, but a gene test for Fertility1 is available from LIC:
http://www.lic.co.nz/lic_News_Archive.cfm?nid=500
Holstein haplotypes for dominant red (HDR) and black/red (HBR)
By Paul VanRaden, Dan Null, Tom Lawlor, and Ben Dorshorst
Red hair color has complex inheritance in Holsteins and is controlled by a new, dominant mutation on chromosome 3 and also by 4 different alleles at the MC1R gene on chromosome 18 (Lawlor et al., 2014). Thus, breeders require more information to track and predict color. Two new haplotype tests that track dominant red (HDR) and black/red (HBR) are reported along with the recessive red haplotype test (HHR) reported since August 2013. Exact gene tests are becoming available, but haplotype tests remain useful for animals genotyped with chips designed before causative alleles were discovered. Animals descending from Rosabel, the source of the dominant red mutation, are processed separately to determine their inheritance for HDR. To avoid problems with animals misreported as red instead of black/red, a subsequent genetic analysis isolates the black allele from all three others (black/red, E+, and e). Then, HBR is separated by haplotype analysis from the E+ and e alleles that are included in HHR. Currently some bulls known to carry black/red are listed as red carriers because they inherited only a portion of the black/red haplotype, and these might be corrected by including black/red as another direct gene test in the future. The E+ and e alleles both cause the same recessive red color, but the laboratory gene tests for E+ and e differ. Caution is advised when using HDR, HBR, and HHR because the complex inheritance of red color may reduce accuracy for these haplotype tests compared to others.
Incorrect input data for MOFRA007045598076 PLUMITIF
By Duane Norman and Paul VanRaden
Montbeliarde bull MOFRA007045598076 PLUMITIF had incorrect data in the USA database in August which was distributed in the multi-trait across-country evaluation (MACE) but was suppressed from the USA evaluation after it was discovered. PLUMITIF had 175 daughters in April and 576 daughters in August, but only about 100 of his 401 added daughters were in first lactation. The other 300 were older daughters that had their pedigrees corrected, but only daughters that are still alive were sent from the processing center and not those that were culled. This caused the productive life (PL) of PLUMITIF to become extreme, nearly twice higher than other bulls and also caused upward bias in other traits because the culled daughters were missing. Because the bias affected all of his daughters, the evaluations for PLUMITIF and his daughters were set back to their April values and the new daughters of PLUMITIF were not distributed. More complete historical data should be available in the December evaluation.