Table 2.
Rank correlations between observed and predicted phenotypes (±SE) and approximate accuracies for lines W1 and WB
| Model | Time-dependent | Rank correlation | Approximate accuracy | ||||
|---|---|---|---|---|---|---|---|
| W1 | % Improvementa | WB | % Improvementa | W1 | WB | ||
| STM | – | 0.135 ± 0.012 | – | 0.170 ± 0.012 | – | 0.44 | 0.46 |
| RMM.t | No | 0.148 ± 0.012 | +10 | 0.185 ± 0.012 | +9 | 0.48 | 0.51 |
| RMM.p | No | 0.162 ± 0.012 | +20 | 0.174 ± 0.012 | +2 | 0.53 | 0.47 |
| GLMM | No | 0.150 ± 0.012 | +11 | 0.190 ± 0.012 | +12 | 0.49 | 0.52 |
| RMM.t | Yes | 0.063 ± 0.013 | −53 | 0.134 ± 0.012 | −21 | 0.20 | 0.37 |
| RMM.p | Yes | 0.049 ± 0.013 | −64 | 0.124 ± 0.012 | −27 | 0.16 | 0.34 |
| GLMM | Yes | 0.081 ± 0.013 | −41 | 0.149 ± 0.012 | −12 | 0.26 | 0.41 |
Rank correlations between observed and predicted phenotypes and approximate accuracies using STM, RMM.t, RMM.p, and GLMM are provided for lines W1 and WB. All models were analysed by either excluding (time-dependent = no) or including (time-dependent = yes) timing of IGE expression
aCompared to STM