Skip to main content
ARS Home » Plains Area » Clay Center, Nebraska » U.S. Meat Animal Research Center » Genetics and Animal Breeding » Research » Publications at this Location » Publication #87374
Title: PROPORTION OF VARIANCE OF CYTOPLASMIC EFFECTS DUE TO REML BIAS FOR MILK YIELD USING ANIMAL AND SIRE MODELS

Author
item RORATO, P - UNIV. OF NEBRASKA-LINCOLN
item Van Vleck, Lloyd
item KEOWN, JEFFREY - UNIV. OF NEBRASKA-LINCOLN

Submitted to: Journal of Animal Science Supplement
Publication Type: Abstract Only
Publication Acceptance Date: 11/17/1997
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Milk yields (2x, 305 d, ME) from first three lactations of 139,869 Holstein cows were used. Cytoplasmic line was assigned by tracing female paths to last female ancestor using DHI records. Data were divided randomly by herd into 10 samples with averages of 13,987 lactations of 6,806 cows with 2,026 cytoplasmic lines. Two single trait models were used with fixed effects of herd-year-seasons of freshening to estimate relative variance due to rando effects of genetic value of cow (g2) (or sire, .25 g2), sire by herd interaction (s2), cytoplasmic line (c2), cow permanent environmental (p2), and residual (e2) for both an animal and unrelated sire model. For each sample, 10 additional analyses were performed with levels for cytoplasmic and sire by herd effects assigned randomly to records in a herd. Estimates were obtained with a derivative-free REML algorithm using numerator relationships computed from samples (animal model). Standard errors (SE) were computed from average information matrix. Mean estimates were compare for 10 original samples (DAT) with 100 simulated analyses (SIM). For animal model, g2 averaged .300 and .321 with sample SE of .025 to.029; p2 averaged .242 and .235 with SE, .024; s2 averaged .015 and .003 with SE, .008; and c2 averaged .011 and .003 with SE, .007 for DAT and SIM, respectively. For the sire model, g2 averaged .232 and .258; s2 averaged .018 and .004, and c2 averaged .038 and .003 for DAT and SIM, respectively, with similar SE as for animal model. Results show REML bias did not account for all of estimates of c2 and s2. In analyses with randomly assigned levels of cytoplasmic lines and sire by herd interactions mean estimates of c2 and s2 were about .003 compared to estimates of .011 to .038 for actual data. Failure to account for s2 and c2 inflated slightly estimates of g2.