QSAR Study of 2-benzylindole Derivatives Using Artificial Neural Network

The quantitative structure-activity relationship(QSAR) of 2-phenylindole derivatives was studied using artificial neural network(ANN) method.The affinities of 36 2-phenylindole derivatives on estrogen receptor in calf uterine tissue were used as activity data.The strutral parameters were the net charges Q of the atoms connected with nitrogen atom,dihedral angle D among C2,C3,C10,C15,bond length L between C5 and the atom joins to C5,and zero grade connection index 0G.The QSAR model of 30 compounds as the training set was constructed based on the improved backpropagation(BP) neural network algorithm.The correlation coefficient was R=0.999 3.The residual 5 compounds was used as the prediction set and the correlation coeffcient was R=0.901 3.The result shows that the fitted performance of ANN method was comparatively precise and the predicted effect was preferable.