Ramasamy Thilagavathi, Selvam Chelliah
Department of Biotechnology, Faculty of Engineering, Karpagam University, Coimbatore, India.
Department of Pharmaceutical Sciences, College of Pharmacy and Health Sciences, Texas Southern University, Houston, TX 77004, USA.
Bioorg Med Chem Lett. 2015 Oct 15;25(20):4632-6. doi: 10.1016/j.bmcl.2015.08.040. Epub 2015 Aug 20.
Virtual screening has become an important tool in drug discovery process. Structure based and ligand based approaches are generally used in virtual screening process. To date, several benchmark sets for evaluating the performance of the virtual screening tool are available. In this study, our aim is to compare the performance of both structure based and ligand based virtual screening methods. Ten anti-cancer targets and their corresponding benchmark sets from 'Demanding Evaluation Kits for Objective In silico Screening' (DEKOIS) library were selected. X-ray crystal structures of protein-ligand complexes were selected based on their resolution. Openeye tools such as FRED, vROCS were used and the results were carefully analyzed. At EF1%, vROCS produced better results but at EF5% and EF10%, both FRED and ROCS produced almost similar results. It was noticed that the enrichment factor values were decreased while going from EF1% to EF5% and EF10% in many cases.
虚拟筛选已成为药物发现过程中的一项重要工具。基于结构和基于配体的方法通常用于虚拟筛选过程。迄今为止，有几个用于评估虚拟筛选工具性能的基准集。在本研究中，我们的目的是比较基于结构和基于配体的虚拟筛选方法的性能。从“客观计算机筛选的严格评估试剂盒”（DEKOIS）库中选择了10个抗癌靶点及其相应的基准集。基于蛋白质-配体复合物的X射线晶体结构的分辨率进行选择。使用了诸如FRED、vROCS等OpenEye工具，并对结果进行了仔细分析。在EF1%时，vROCS产生了更好的结果，但在EF5%和EF10%时，FRED和ROCS都产生了几乎相似的结果。注意到在许多情况下，从EF1%到EF5%和EF10%时，富集因子值会降低。
