부산대학교 도서관

Molecular docking simulation is a useful tool in the prediction of protein-ligand binding affinity on alarge scale and has great potential in various applicationfields such as virtual screening of potentialdrug molecules. However, the reliability of molecular docking is still weak in the estimation of ligand-binding free energy, which limits the applicability of molecular docking simulation. Ligand torsionnumber is related to theflexibility of ligand and generally incorporated as a crucial variable in thethermodynamic function of binding free energy. In this study, we investigated how the ligand torsionnumber has influence on the binding affinity prediction of AutoDock, a popular molecular dockingsimulation tool. The pKd values of various protein-ligands were estimated by using the binding freeenergy function of AutoDock and compared with their experimental pKd values. The torsion numberdependent comparison showed that the predicted binding affinities were mostly underestimated inthe complexes of higher torsion numbers, whereas the underestimated and overestimated cases wererelatively balanced at relatively lower torsion numbers. A new weight factor for torsion-free energyterm of binding energy function was determined and introduced to make correction to theunderestimation of binding affinity of ligands with high torsion numbers. It is expected that the torsionnumber dependent deviation pattern of AutoDock and its correction strategy are useful in the large-scale validation of protein-ligand binding affinity.