Centro de Bioinformática y Simulación Molecular, Facultad de Ingeniería en Bioinformática, Universidad de Talca, 2 Norte 685, Casilla 721, Talca, Chile.
J Mol Graph Model. 2010 Nov;29(3):363-71. doi: 10.1016/j.jmgm.2010.08.005. Epub 2010 Sep 21.
Inhibitory activities of flavonoid derivatives against aldose reductase (AR) enzyme were modelled by using CoMFA, CoMSIA and GALAHAD methods. CoMFA and CoMSIA methods were used for deriving quantitative structure-activity relationship (QSAR) models. All QSAR models were trained with 55 compounds, after which they were evaluated for predictive ability with additional 14 compounds. The best CoMFA model included both steric and electrostatic fields, meanwhile, the best CoMSIA model included steric, hydrophobic and H-bond acceptor fields. These models had a good predictive quality according to both internal and external validation criteria. On the other hand, GALAHAD was used for deriving a 3D pharmacophore model. Twelve active compounds were used for deriving this model. The obtained model included hydrophobe, hydrogen bond acceptor and hydrogen bond donor features; it was able to identify the active AR inhibitors from the remaining compounds. These in silico tools might be useful in the rational design of new AR inhibitors.
通过使用 CoMFA、CoMSIA 和 GALAHAD 方法，对黄酮衍生物对醛糖还原酶 (AR) 酶的抑制活性进行了建模。CoMFA 和 CoMSIA 方法用于推导定量构效关系 (QSAR) 模型。所有 QSAR 模型均使用 55 种化合物进行训练，然后使用另外 14 种化合物对其进行预测能力评估。最佳 CoMFA 模型同时包含立体和静电场，而最佳 CoMSIA 模型则包含立体、疏水和氢键接受体场。根据内部和外部验证标准，这些模型具有良好的预测质量。另一方面，GALAHAD 用于推导 3D 药效团模型。使用 12 种活性化合物来推导该模型。得到的模型包含疏水、氢键接受体和氢键供体特征；它能够从其余化合物中识别出活性的 AR 抑制剂。这些计算机工具可能有助于合理设计新的 AR 抑制剂。
Zotero 插件