Skip to main content
. 2019 May 6;14(5):e0208871. doi: 10.1371/journal.pone.0208871

Table 3. Analysis of factors that influence the variation in predictive ability of single and multi-environment prediction models.

The effects of the statistical model (GBLUP, RKHS-1 and RKHS-2), the trait (DTF, GY and PH), the cross-validation strategy (CV1 and CV2), drought stress environments (E1, E2 and E3) and different combinations of training environment and their interactions were evaluated.

Analysis with only main effects
Source DF SS MS F-value ProbF
Model 10 309.728 30.973 1 773.667 < 0.0001
Error 32 389 565.595 0.017
Corrected Total 32 399 875.323
Trait 2 145.581 72.791 7 141.750 < .0001
Model 1 10.730 10.730 1 052.770 < .0001
Method 2 1.426 0.713 69.930 < .0001
CV 1 15.993 15.993 1 569.160 < .0001
Drought environment 2 134.409 67.204 6 593.650 < .0001
Training environment 2 1.588 0.794 77.880 < .0001
Analysis with main effects and all first-order interactions
Model 51 496.946 9.744 833.032 < 0.0001
Error 32 348 378.377 0.012
Corrected Total 32 399 875.323
Trait 2 145.581 72.791 7 141.750 < .0001
Model 1 10.730 10.730 1 052.770 < .0001
Method 2 1.426 0.713 69.930 < .0001
CV 1 15.993 15.993 1 569.160 < .0001
Drought environment 2 134.409 67.204 6 593.650 < .0001
Training environment 2 1.588 0.794 77.880 < .0001
Trait × Model 2 4.605 2.303 225.920 < .0001
Trait × Method 4 1.418 0.355 34.780 < .0001
Model × Method 2 1.812 0.906 88.910 < .0001
Trait × CV 2 1.882 0.941 92.330 < .0001
Model × CV 1 15.993 15.993 1 569.160 < .0001
Method × CV 2 1.627 0.814 79.820 < .0001
Trait × Drought 4 152.543 38.136 3 741.620 < .0001
Model × Drought 2 0.786 0.393 38.580 < .0001
Method × Drought 4 0.499 0.125 12.240 < .0001
CV × Drought 2 0.792 0.396 38.840 < .0001
Trait × Training environment 4 0.957 0.239 23.470 < .0001
Model × Training environment 2 1.588 0.794 77.880 < .0001
Method × Training environment 4 0.935 0.234 22.930 < .0001
CV × Training environment 2 1.782 0.891 87.440 < .0001
Drought × Training environment 4 1.686 0.422 41.350 < .0001

DF: Degree of freedom; SS: Sum of squares; MS: Mean square; ×: interaction between the factors.