QSAR and Molecular Docking Studies for the Inhibition of MAPK Interacting Kinase

The MAPK-interacting kinases (MNK1/2) play an important role in cancer treatment. Accordingly, inhibition of MNK1 can be a significant target for developing safe and nontoxic treatment methods for cancer. In this work, a set of MNK1 inhibitors are analyzed as anticancer agents using quantitative structure-activity relationship (QSAR) models. The meta-heuristic optimization algorithm genetic algorithm (GA) is first used to select accurate descriptors for constructing the QSAR models. The Multiple Linear Regression (MLR) method has been used to validate the models. The applications of the techniques like Back Propagation Neural Network (BPNN) and Support Vector Machine (SVM) provides added advantage in the prediction of the desired end point. The development of the therapeutic activities of 47 different inhibitors against MNK1 has been analyzed through molecular docking. The docking process helped to estimate the binding energy, ligand-protein interaction to evaluate the inhibition-potential. Leave-one-out cross-validation and Y-scrambling have been used to verify the robustness of the QSAR models. A high correlation is discovered between predicted and experimental values which confirm the fine quality of the QSAR models. The molecular docking results showed that the MNK1 receptor was well inhibited by the studied ligands showing the least binding energy (-8.2 Kcal/mol). A Review of the statistical measures obtained from the results of BPNN and SVM has discovered that the models predicted by GA-MLR are sufficiently reliable.