Test set of compounds

In order to estimate the true predictive power of a QSAR model, one needs to compare the predicted and observed activities of a sufficiently large external test set of compounds that were not used in the model development. One convenient parameter is an external defined as follows (similar to Eq. (1) for the training set) ... 

To check the reliability of the model and to assess its predictive power, a validation step was performed by predicting affinity data for a large test set of compounds collected from three different sources, (i) Glutarimido derivatives belonging to the same class of compounds used to build the training set were evaluated against the pharmacophore model to predicting their affinity values. 

Test Set 1. A more realistic trial was the selection, as a test set, of compounds as they had been input into the program. 

Figure 9.6 The hypothetical relationship between distance to the model space/domain of applicability and prediction error for a test set of compounds.

Once the classificator (wg and the w,) is known new compounds can be classified by inserting their values of x, into equation (45). Reclassification of the compounds of the training series is performed in the same way. The separating power of the classificator is judged in the usual way by the error of reclassification and the error of classification for a test set of compounds or for a simulated prediction using the leave- -out technique as already discussed in eoimection with discriminant analysis. In two-class problems sign and magnitude... 

Calculations of proton chemical shifts that used several common density functionals and basis sets on a test set of compounds together with Unear scaling of the results recommended WP04/cc-pVDZ(rms error 0.115 ppm, m = 1.020, ft = 31.84), WP04/6-31G (rms error 0.12 ppm, m = 1.033, ft = 32.02), B3LYP/6-31G (rms error 0.13 ppm. 

---