Development linear and non-linear QSAR models for predicting AXL kinase inhibitory activity of N-[4-(quinolin-4-yloxy)phenyl]benzenesulfonamides

Abstract

In this research, we used CoMFA, LSSVM and FFANN for creating QSAR models for predicting AXL Kinase inhibitory activity of N-[4-(Quinolin-4-yloxy)phenyl]benzenesulfonamides. A CoMFA model with three components was developed and CoMFA contour maps were interpreted to extract chemical features that influence the inhibitory activity of these molecules. R² for train and test set of CoMFA model were 0.8900 and 0.8171, respectively. Model created by five Dragon descriptors and LSSVM model showed slightly better predictive power with respect to CoMFA model. R² for train, test set of created LSSVM model were 0.0.8477 and 0.8218, respectively. Also, a FFANN model, using the same five descriptors, was developed with 2 neurons in its hidden layer and R² for its train and test sets were 0.8314 and 0.8522, respectively. All created models were validated by calculating several statistical parameters and their applicability domain were investigated by calculating leverage. Furthermore, a homology model was built for Axl structure and molecules with the lowest and the greatest activity were docked to it and their interactions with Axl were investigated.

Keywords:

• QSAR
• Molecular docking
• Axl receptor
• Cancer
• N-[4-(Quinolin-4-yloxy)phenyl]benzenesulfonamides

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.