Correlation analysis. Its application to the structure-activity relation of triazines inhibiting dihydrofolate reductase

A new procedure for improving the predictiveness of CoMFA models and its application to a set of dihydrofolate reductase inhibitors

“…15. test molecules are predicted after a fit to the structures in the training set. As an example, the classical QSAR data set of Hansch and co-workers was used [14]. From this ensemble of 256 dihydrofolate reductase inhibitors, two training sets consisting of 80 structures each and a test set of 70 compounds were randomly chosen.…”

Neighborhood Behavior: Validation of Two‐Dimensional Molecular Similarity as a Predictor of Similar Biological Activities and Docking Scores

“…The mean pairwise dissimilarity values for the two largest datasets [26,27] recalculated by Accelrys Accord software were 0.09 and 0.08 only. Dixon and Merz [9] used the same datasets as Patterson et al Matter [10] has studied several larger datasets: the dataset of 256 triazine inhibitors of dihydrofolate reductase [28] is the largest, with a recalculated mean pairwise dissimilarity value of 0.21. Only in a recent series of papers by Horvath and co-workers [11 -13] notably larger datasets have been used.…”

Quantitative structure-activity relationships by neural networks and inductive logic programming. II. The inhibition of dihydrofolate reductase by triazines

“…The data in this study, 186 2,4-diamino-6,6-dimethyl-5-phenyl-dihydrotriazines (I) ( Table 1) are a subset of those used to derive a QSAR by multilinear regression [2] and by neural networks [3]. The 186 triazines haTce two or three rings, with the phenyl ring (the second ring) substituted at either the 3-or 4-position; ortho-substituents were not considered as there were only 11 examples, which is too small a set for generalisations [4].…”

“…Silipo and Hansch [2] correlated the activity of the triazines with the hydrophobic parameters and the molar refractivities MR of the 3-and 4-substituents of the phenyl ring. The activity was measured by log l/C, where C is the molar concentration that produces 50% reversible inhibition of dihydrofolate reductase, obtained from L1210 mouse leukaemia cells and Walker 256 rat tumours.…”

“…The data come from work of Baker and co-workers in the 1960s, who synthesised many triazines and studied their inhibition of dihydrofolate reductase (DHFR). Silipo and Hansch [2] concluded that the data were 'an excellent testing ground for further new approaches to structure-activity analysis'.…”