Basic Medical Science and Molecular Medicine, Tokai University School of Medicine, Isehara, Kanagawa 259-1143, Japan.
J Comput Chem. 2010 Aug;31(11):2225-32. doi: 10.1002/jcc.21517.
Alpha sphere filter (ASF) method is a novel previrtual screening method to undertake a rapid virtual screening of a huge chemical space. The small-molecule binding site of a target molecule can be characterized by a set of alpha spheres generated at the site. Two types of pseudomolecules representing molecules that likely fit into the binding site were molded from the set of alpha spheres. Based on the pseudomolecules, pseudomolecular descriptors corresponding to the conventional two-dimensional (2D) molecular descriptors were calculated. The correlations between the pseudomolecular descriptors and the 2D molecular descriptors were analyzed for a set of high-quality X-ray structures of the complexes between ligands and proteins. By use of these correlations, specific value ranges of the 2D molecular descriptors were determined. These value ranges were applied in virtual screening. In a trial to screen 200 active ligands out of a chemical database with 42,547 molecules, the enrichment rate of 5.8 was attained. The enrichment rate was good enough for a prescreening tool prior to docking simulations. As the ASF method screens molecules by 2D molecular descriptors, it is rapid enough to screen a huge chemical space and could significantly decrease the number of trivial compounds to be considered in the following docking simulations. Therefore, the ASF method can contribute to enlarge the possibility of virtual screening.
阿尔法球筛选(ASF)方法是一种新颖的预虚拟筛选方法,可快速对庞大的化学空间进行虚拟筛选。目标分子的小分子结合位点可以通过在该位点生成的一组阿尔法球来描述。从一组阿尔法球中塑造了两种代表可能适合结合位点的分子的伪分子。基于伪分子,计算了与传统二维(2D)分子描述符相对应的伪分子描述符。对配体与蛋白质之间复合物的一组高质量 X 射线结构进行了 2D 分子描述符与伪分子描述符之间相关性的分析。利用这些相关性,确定了 2D 分子描述符的特定值范围。这些值范围应用于虚拟筛选。在对 42547 个分子的化学数据库中筛选 200 种活性配体的试验中,获得了 5.8 的富集率。对于对接模拟之前的预筛选工具来说,这个富集率已经足够好了。由于 ASF 方法通过 2D 分子描述符筛选分子,因此它足够快速,可以筛选庞大的化学空间,并可以显著减少后续对接模拟中需要考虑的琐碎化合物的数量。因此,ASF 方法可以有助于增加虚拟筛选的可能性。