Table 1. Average correlations between predicted genetic values and observed phenotypes for different CV procedures with GBLUP and different traits.
type of CV | starvation resistance | startle response |
(4∶1)-CV1 all2 | 0.2393 (0.008) | 0.230 (0.012) |
(3∶2)-CV all | 0.213 (0.006) | 0.216 (0.011) |
(2∶3)-CV all | 0.176 (0.006) | 0.181 (0.010) |
(1∶4)-CV all | 0.124 (0.006) | 0.128 (0.006) |
(4∶1)-CV male - female4 | 0.164 (0.007) | 0.217 (0.011) |
(4∶1)-CV female - male | 0.182 (0.007) | 0.235 (0.012) |
(4∶1)-CV male - male | 0.203 (0.008) | 0.230 (0.012) |
(4∶1)-CV female - female | 0.254 (0.009) | 0.216 (0.011) |
“-CV” means: parts are used as training set and parts are used as validation set.
The average of the medians of male and female measurements was used to predict line phenotypes. Predicted phenotypes were then correlated with the averages of the medians of male and female measurements.
Average correlation between predicted genetic values and observed phenotypes. Results are averages over 20 replicates. Standard errors of the means in parentheses.
“CV sex sex” means: Medians of measurements of sex were used in the training set, medians of sex were used in the validation set.