Structure-based and multiple potential three-dimensional quantitative structure-activity relationship (SB-MP-3D-QSAR) for inhibitor design.
Qi-Shi Du,Jing Gao,Yu-Tuo Wei,Li-Qin Du,Shu-Qing Wang,Ri-Bo Huang
DOI: https://doi.org/10.1021/ci300066y
IF: 6.162
2012-01-01
Journal of Chemical Information and Modeling
Abstract:The inhibitions of enzymes (proteins) are determined by the binding interactions between ligands and targeting proteins. However, traditional QSAR (quantitative structure-activity relationship) is a one-side technique, only considering the structures and physicochemical properties of inhibitors. In this study, the structure-based and multiple potential three-dimensional quantitative structure-activity relationship (SB-MP-3D-QSAR) is presented, in which the structural information of host protein is involved in the QSAR calculations. The SB-MP-3D-QSAR actually is a combinational method of docking approach and QSAR technique. Multiple docking calculations are performed first between the host protein and ligand molecules in a training set. In the targeting protein, the functional residues are selected, which make the major contribution to the binding free energy. The binding free energy between ligand and targeting protein is the summation of multiple potential energies, including van der Waals energy, electrostatic energy, hydrophobic energy, and hydrogen-bond energy, and may include nonthermodynamic factors. In the foundational QSAR equation, two sets of weighting coefficients {aj} and {bp} are assigned to the potential energy terms and to the functional residues, respectively. The two coefficient sets are solved by using iterative double least-squares (IDLS) technique in the training set. Then, the two sets of weighting coefficients are used to predict the bioactivities of inquired ligands. In an application example, the new developed method obtained much better results than that of docking calculations.