School of Pharmacy, East China University of Science and Technology, 130 Meilong Road, Shanghai 200237, China.
J Mol Model. 2010 Dec;16(12):1867-76. doi: 10.1007/s00894-010-0690-z. Epub 2010 Mar 12.
Three-dimensional pharmacophore models of human adenosine receptor A(₂A) antagonists were developed based on 23 diverse compounds selected from a large number of A(₂A) antagonists. The best pharmacophore model, Hypo1, contained five features: one hydrogen bond donor , three hydrophobic points and one ring aromatic. Its correlation coefficient, root mean square deviation, and cost difference values were 0.955, 0.921 and 84.4, respectively, suggested that the Hypo1 model was reasonable and reliable. This model was validated by three methods: a test set of 106 diverse compounds, a simulated virtual screening, and superimposition with the crystal structure of A(₂A) receptor. The results showed that Hypo1 was not only in agreement with the A(₂A) crystal structure and literature reports, but also well identified active A(₂A) antagonists from the virtual database. This methodology provides helpful information and a robust tool for the discovery of potent A(₂A) antagonists.
基于从大量 A₂A 拮抗剂中选择的 23 种不同化合物,开发了人源腺苷受体 A₂A 拮抗剂的三维药效团模型。最佳药效团模型 Hypo1 包含五个特征:一个氢键供体、三个疏水性点和一个环芳基。其相关系数、均方根偏差和成本差异值分别为 0.955、0.921 和 84.4,表明 Hypo1 模型合理可靠。该模型通过三种方法进行验证:一组 106 种不同化合物的测试集、模拟虚拟筛选以及与 A₂A 受体的晶体结构叠加。结果表明,Hypo1 不仅与 A₂A 晶体结构和文献报道一致,而且还能很好地从虚拟数据库中识别出活性 A₂A 拮抗剂。该方法为发现有效的 A₂A 拮抗剂提供了有价值的信息和强大的工具。