Table 4.
Mean (n = 10) cross-validation predictive ability in a multi-population (MP) training set using three different SNP sets for seven HA traits and Comb DTH
| Trait | RF | GBLUP | RR | KGD | |||||
|---|---|---|---|---|---|---|---|---|---|
| Set 1 | Set 2 | Set 3 | Set 1 | Set 2 | Set 3 | Set 1 | Set 2 | Set 3 | |
| Aor STD | 0.23 (0.007) | 0.23 (0.008) | 0.23 (0.003) | 0.23 (0.006) | 0.22 (0.006) | 0.21 (0.006) | 0.22 (0.009) | 0.19 (0.007) | 0.22 (0.005) |
| Aor SEV | 0.26 (0.004) | 0.26 (0.004) | 0.24 (0.005) | 0.24 (0.004) | 0.24 (0.005) | 0.23 (0.006) | 0.21 (0.005) | 0.23 (0.005) | 0.24 (0.004) |
| Lin STD | 0.11 (0.006) | 0.14 (0.005) | 0.15 (0.010) | 0.10 (0.003) | 0.10 (0.003) | 0.07 (0.009) | 0.13 (0.006) | 0.15 (0.006) | 0.10 (0.011) |
| Rua STD | 0.26 (0.007) | 0.25 (0.007) | 0.25 (0.004) | 0.26 (0.005) | 0.27 (0.006) | 0.26 (0.005) | 0.28 (0.006) | 0.30 (0.007) | 0.26 (0.005) |
| Rua SEV | 0.32 (0.008) | 0.31 (0.006) | 0.30 (0.006) | 0.31 (0.006) | 0.31 (0.005) | 0.31 (0.005) | 0.28 (0.010) | 0.30 (0.008) | 0.30 (0.007) |
| Rua STD + SEV | 0.40 (0.006) | 0.40 (0.005) | 0.39 (0.004) | 0.43 (0.006) | 0.44 (0.007) | 0.44 (0.007) | 0.44 (0.007) | 0.44 (0.008) | 0.43 (0.006) |
| Comb SEV | 0.36 (0.007) | 0.35 (0.006) | 0.34 (0.005) | 0.35 (0.007) | 0.36 (0.007) | 0.36 (0.007) | 0.33 (0.011) | 0.35 (0.010) | 0.36 (0.006) |
| Comb DTH | 0.47 (0.004) | 0.47 (0.004) | 0.46 (0.005) | 0.52 (0.002) | 0.52 (0.003) | 0.52 (0.003) | 0.51 (0.002) | 0.51 (0.003) | 0.52 (0.005) |
Set 1 = 43,966 SNPs (1% missing data per SNP site). Set 2 = 249,546 SNPs (10% missing data per SNP site). Set 3 = 1,023,011 SNPs (50% missing data per SNP site)
RR ridge regression, RF random forest, KGD GBLUP using KGD genomic relationship matrix