By taking advantage of the high-performance computers, docking-based virtual screening (VS) has gained more and more attention, and has become one of the core technologies in drug design and development. However, the prediction accuracy of molecular docking may be impaired by some inherent defects, such as simplified scoring functions and ignorance of protein flexibility. In this talk, I will discuss some new strategies developed in my group to improve the efficiency and accuracy of docking-based virtual screening. In the first part, I will discuss the MIEC-SVM approach based on free energy decomposition and machine learning algorithm, which shows good capability to identify binding peptides of modular domains and small molecule inhibitors of drug targets. In the second part, I will discuss a novel parallel virtual screening strategy by integrating molecular docking and complex-based pharmacophore searching based on multiple protein structures.
浙江大学药学院求是特聘教授，药物信息和计算生物学中心主任。长期围绕计算机辅助药物设计中的核心问题展开前沿交叉学科研究，在Chemical Reviews, ACS Central Science, Nucleic Acids Research, PNAS, Briefings in Bioinformatics、Journal of Medicinal Chemistry, Journal of Chemical Theory and Computation, Journal of Chemical Information and Modeling等知名期刊发表SCI论文350余篇，SCI引用12000余次，H因子60；获授权专利和软件著作权28项。任中国化学会计算(机)化学专业委员会副主任委员兼秘书长，J Cheminf、J Chem Inf Model、Int J Mol Sci等14种SCI期刊编委或顾问编委。