论文部分内容阅读
目的:构建和验证GPR40激动剂药效团模型。方法:采用生物活性跨越1~160000 nmol?L~(-1)的23个GPR40激动剂小分子作为训练集进行药效团模型的构建;利用测试集验证、Fisher验证及分子对接方法对模型进行评价;根据所构建的药效团模型,设计并合成2个全新结构的化合物,通过正常小鼠糖耐量试验的验证。结果:筛选得到由1个氢键受体(HBA)、2个疏水中心(HP)、1个芳环中心(RA)组成的最佳模型(Correl.=0.971,Config=16.645,△cost=58.370),依据药效团模型设计得到的与药效团匹配较高的化合物Lyb-438能显著降低正常小鼠餐后0.5 h血糖(P<0.05)。结论:所构建的药效团模型具有较强的预测能力和较高的可信性,为进一步的数据库搜索及寻找新型小分子GPR40激动剂提供了依据。
Objective: To construct and validate GPR40 agonist pharmacophore model. Methods: The pharmacophore model was constructed with 23 small GPR40 agonists whose biological activity spanned 1 ~ 160000 nmol? L ~ (-1) as a training set. Using the test set validation, Fisher validation and molecular docking methods, According to the constructed pharmacophore model, two novel compounds were designed and synthesized, which were verified by normal mice glucose tolerance test. Results: The best model (Correl. = 0.971, Config = 16.645, △ cost = 58.370) was obtained by screening one HBA, two hydrophobic centers and one aromatic center (RA) ), And Lyb-438, a compound matching pharmacophore, was designed according to pharmacophore model to reduce the postprandial blood glucose 0.5 h (P <0.05). CONCLUSION: The constructed pharmacophore model has strong predictive ability and high credibility, which provides a basis for further database search and searching for new small molecule GPR40 agonists.