TABLE 2.
Accuracy and bias of prediction using different models through 20 replicates of fivefold cross‐validation in four populations
| Species (heritability (SE)a) | Method | Accuracy | Regression coefficient |
|---|---|---|---|
| Atlantic salmon (0.25 (0.06)) | BLUP | 0.510 (0.106) | 1.012 (0.281) |
| GBLUP | 0.615 (0.101) | 1.019 (0.224) | |
| BayesR | 0.611 (0.102) | 1.054 (0.259) | |
| WGBLUP | 0.627 (0.101) | 0.900 (0.243) | |
| GFBLUP | 0.560 (0.102) | 0.513 (0.107) | |
| Common carp (0.26 (0.06)) | BLUP | 0.591 (0.113) | 0.980 (0.239) |
| GBLUP | 0.635 (0.125) | 1.046 (0.241) | |
| BayesR | 0.747 (0.124) | 0.994 (0.200) | |
| WGBLUP | 0.657 (0.114) | 0.892 (0.259) | |
| GFBLUP | 0.540 (0.129) | 0.478 (0.119) | |
| Sea bream (0.12 (0.06)) | BLUP | 0.462 (0.197) | 1.243 (0.816) |
| GBLUP | 0.625 (0.204) | 1.153 (0.586) | |
| BayesR | 0.643 (0.206) | 1.906 (1.017) | |
| WGBLUP | 0.636 (0.206) | 0.960 (0.524) | |
| GFBLUP | 0.574 (0.193) | 0.382 (0.143) | |
| Rainbow trout (0.50 (0.06)*) | BLUP | NA | NA |
| GBLUP | 0.816 (0.079) | 0.992 (0.126) | |
| BayesR | 0.829 (0.072) | 0.978 (0.118) | |
| WGBLUP | 0.831 (0.076) | 0.929 (0.123) | |
| GFBLUP | 0.771 (0.082) | 0.730 (0.087) |
Standard deviations in brackets for accuracy and regression coefficient. NA: The BLUP method was not available because of the lack of pedigree in the rainbow trout population.
Abbreviations: BayesR, Bayesian mixture model; BLUP, BLUP method based on pedigree; GBLUP, genomic BLUP; GFBLUP, genomic feature BLUP with GWAS p‐value of 0.05 as genomic feature; WGBLUP, weighted GBLUP.
a
Heritability (standard error, SE) estimated using pedigree relationship information, except for rainbow trout (* heritability estimated using genomic relationship information).