QUANTITATIVE STRUCTURE-ACTIVITY-RELATIONSHIPS BY NEURAL NETWORKS AND INDUCTIVE LOGIC PROGRAMMING .2. THE INHIBITION OF DIHYDROFOLATE-REDUCTASE BY TRIAZINES

One of the largest available data sets for developing a quantitative s tructure-activity relationship (QSAR) - the inhibition of dihydrofolat e reductase (DHFR) by 2,4-diamino-6,6-dimethyl-5-phenyl-dihydrotriazin e derivatives - has been used for a sixfold cross-validation trial of neural networks, inductive logic programming (ILP) and linear regressi on. No statistically significant difference was found between the pred ictive capabilities of the methods. However, the representation of mol ecules by attributes, which is integral to the ILP approach, provides understandable rules about drug-receptor interactions.