The discovery of potential cyclin A/CDK2 inhibitors: a combination of 3D QSAR pharmacophore modeling, virtual screening, and molecular docking studies
The discovery of potential cyclin A/CDK2 inhibitors: a combination of 3D QSAR pharmacophore modeling, virtual screening, and molecular docking studies
复制标题
DOI：
10.1007/s00044-013-0571-y
复制
发表时间：
2013-12-01
影响因子：
2.6
通讯作者：
Sevin, Fatma
中科院分区：
文献类型：
--
作者：
Ece, Abdulilah;Sevin, Fatma
细胞周期蛋白依赖性激酶是调节细胞周期过程的酶家族。它们已被发现是潜在抗癌药物的新靶点。在本研究中，一个3D药效团模型已开发的细胞周期蛋白A/CDK 2从其已知的抑制剂。最可靠的定量HypoGen模型（Hypo 1）由两个氢键受体，一个氢键供体和一个疏水特征组成。相关系数为0.98，均方根偏差为0.84，配置成本为16.25，成本差为102.93的Hypo 1显示出显著的预测能力，并且具有> 90%的概率表示活动数据中的真实相关性。在95%置信水平下使用Fisher检验和检验集预测（r = 0.96）验证模型。然后使用Hypo 1对Life Chemicals和NCI 2003数据库进行虚拟筛选，其中每个化合物产生多种构象（596，030种化合物，45，603，414种构象异构体）。根据Lipinski，Ghose和Veber的规则过滤命中。在对接模拟之后，使用一致性评分来确定与蛋白质结合位点最佳相互作用的配体姿势并减少假阳性的数量。11个命中最终被选为有效的候选线索。这项工作可能有助于基于确定的命中识别或设计新型抗癌药物。在这项研究中获得并验证的药效团模型可以用作在其他化合物数据库中搜索CDK 2抑制剂的三维查询。
Cyclin-dependent kinases are a family of enzymes that regulates the cell cycle process. They have been found to be novel targets for potential anti-cancer drugs. In the present study, a 3D pharmacophore model has been developed for cyclin A/CDK2 from its known inhibitors. The most reliable quantitative HypoGen model (Hypo1) consists of two hydrogen bond acceptors, one hydrogen bond donor and one hydrophobic feature. Hypo1, with a correlation coefficient of 0.98, a root mean square deviation of 0.84, a configuration cost of 16.25 and a cost difference of 102.93, showed a remarkable predictive power and has > 90 % probability of representing a true correlation in the activity data. The model was validated using Fisher's test at 95 % confidence level and test set prediction (r = 0.96). Hypo1 was then employed for virtual screening of Life Chemicals and NCI2003 databases of which multiple conformations were generated for each compound (596,030 compounds, 45,603,414 conformers). Hits were filtered according to the Lipinski, Ghose, and Veber's rules. Following docking simulations, consensus scoring was used to determine the ligand poses that interact best with the protein binding site and to reduce number of false positives. 11 hits were ultimately selected as potent candidate leads. This work may help in the identification or design of novel anti-cancer drugs based on hits determined. The pharmacophore model obtained and validated in this study can be used as a three-dimensional query in searches for CDK2 inhibitors in additional compound databases.