Construction of 3D-QSAR models to predict antiamoebic activities of pyrazoline and dioxazoles derivatives

1-N-substituted thiocarbamoyl-3-phenyl-2-pyrazolines and 3,5-substituted-1,4,2-dioxazoles are potent antiamoebic agents. A 3D-QSAR study is applied to a set of 63 molecules. With The multiple linear regression method (MLR) (r = 0.95), the predicted values of activities are in good agreement with the experimental results. The artificial neural network (ANN) techniques, considering the relevant descriptors obtained from the MLR, showed good results; a correlation coefficient of 0.96 was obtained with an 8-3-1 ANN model. As a result of quantitative structure-activity relationships between 1-N-substituted thiocarbamoyl-3-phenyl-2-pyrazolines and 3,5-substituted-1,4,2-dioxazoles, we found that the model proposed in this study is constituted of major descriptors used to describe these molecules. This model is statistically significant and shows very good stability towards data variation in leave-one-out (LOO) cross- validation (rcv =0.90).