Skip to main content
. 2016 Feb 1;48:8. doi: 10.1186/s12711-016-0186-0

Table 5.

Accuracy of prediction of phenotypes in cross-validation

Breed Model Parameter Trait
CI Milk yield Fat yield Protein yield
Holstein A BV 0.080 (0.008) 0.346 (0.011) 0.270 (0.017) 0.266 (0.018)
A + D BV 0.080 (0.008) 0.347 (0.011) 0.270 (0.017) 0.267 (0.018)
DV 0.025 (0.014) 0.038 (0.014) 0.038 (0.003) 0.035 (0.008)
BV + DV 0.082 (0.011) 0.348 (0.011) 0.273 (0.017) 0.269 (0.018)
A vs. A + D P value (BVA vs. BVA + D) 0.9476 0.0169 0.4517 0.2360
P value (BVA vs. [BV + DV]A + D) 0.5456 0.0694 0.0060 0.0202
Jersey A BV 0.094 (0.007) 0.368 (0.017) 0.243 (0.025) 0.256 (0.025)
A + D BV 0.094 (0.007) 0.367 (0.018) 0.244 (0.026) 0.255 (0.026)
DV 0.005 (0.014) 0.046 (0.006) 0.059 (0.018) 0.071 (0.015)
BV + DV 0.094 (0.007) 0.370 (0.017) 0.250 (0.024) 0.262 (0.023)
A vs. A + D P value (BVA vs. BVA + D) 0.6069 0.8485 0.7820 0.6068
P value (BVA vs. [BV + DV]A + D) 0.6000 0.1275 0.1355 0.1727

Average correlations between estimated genetic effects (breeding values [BV], dominance deviations [DV] or total genetic values [BV + DV]) obtained from models A (additive effects only) and A + D (additive and dominance effects) and adjusted phenotypes for calving interval (CI), milk, fat and protein yield in Holstein and Jersey cows based on fivefold cross validation

Standard errors (SE) in parentheses