Submitted to: Journal of Animal Science Supplement
Publication Type: Abstract Only
Publication Acceptance Date: July 1, 1999
Publication Date: N/A
Technical Abstract: To evaluate the use of genetic markers in breeding value prediction, breeding values for heifer ovulation rate calculated using BLUP with or without covariates for expected inheritance of founding animal alleles at selected positions in the genome (FACOV) were compared to subsequent measurements of ovulation rate. Expected inheritance of founding animal alleles was determined from a multi-locus peeling algorithm using marker genotypes. Averages of six ovulation rate measurements made from 12 to 18 months of age were calculated for 426 heifers born in four calving seasons. Sixteen FACOV were selected for the final analysis based on preliminary analyses using data collected prior to the comparison data. These 16 covariates consisted of five founding animal alleles at one location on one chromosome, six at another distant location on the same chromosome, and five on a second chromosome. Including the 16 FACOV reduced the estimate of polygenic variance by 13%. The regressions of observed ovulation rate on predicted breeding value were not substantially or significantly different from the expected value of 1.0 either with (.93 +/- .19) or without (.92 +/- .21) FACOV. The proportion of observed variation in ovulation accounted for by prediction was .052 with FACOV and .043 without FACOV. Predicted breeding values using FACOV were partitioned into parental average and deviation from parental average. The deviation from parental average accounted for 28% of the variance accounted for by prediction with FACOV. Using marker information to track a small number of founding animal alleles increased the accuracy of predicted breeding values by accounting for some within family segregation.