Abstract

Quantitative structure activity relationship approach using enhanced replacement method for variable selection was applied to a series of imidazole containing tetrahydrobenzodiazepines as inhibitors of farnesyltransferase. For the purpose the dataset was divided into training set of 31 compounds and test set of 5 compounds using k means clustering. Statistically significant equations were obtained with high correlation coefficient coefficient (R=0.9637) and low standard deviation (S=0.3715). The robustness of the model was confirmed with the help of R2cv with a value of 0.8886, Y scrambling and by predicting the activities of test compounds. A good correlation of KierA3 shape indices and polarizability with the farnesyltransferase inhibitory activity was achieved.

Keywords: QSAR, Farnesyltransferase Inhibitors, Tetrahydrobenzodiazepines, Y scrambling, Polarizability, Imidazole, Ras proteins, PFT inhibitors, metalloenzyme, GABAA, Kubinyi function (FIT), apol, Specific Polarizability, Spol(R4)