Parameters for the QSAR model developed for the I942 analoguesa.
| Training setbc | Test setd | Cross-validation (CV) | Threshold36,37 | |
|---|---|---|---|---|
| R 2 | 0.972 ± 0.008 | 0.929 ± 0.021 | 0.772 ± 0.055 | R 2 > 0.600 and > 0.500 for CV |
| σ | 27.69 ± 0.60 | 28.69 ± 2.40 | — | — |
| RMSE | 11.83 ± 1.81 | 19.09 ± 3.13 | 12.78 ± 1.50 | RMSE < σ |
| k | 0.972 ± 0.008 | 0.960 ± 0.090 | — | 0.850 ≤ k ≤ 1.150 |
Standard deviations were computed using data from eleven different partitioning of training vs. test sets.
Number of molecules in the training set was 45.
Number of descriptors were ≤9.
Number of molecules in the test set was 11.