摘要
用分子对接和三维全息原子场作用矢量方法对36个来曲唑类衍生物和34个阿那曲唑类衍生物与芳香化酶的作用模式进行了研究,建立了三维定量构效关系模型,并在分子水平上阐述了其结合机制.运用多元线性回归(MLR)建模,同时采用内部及外部双重验证的办法对所得模型稳定性能进行深入分析和检验.MLR建模的复相关系数(Rcum)、留一法交互校验复相关系数(QCV)和外部样本校验复相关系数(Qext)分别为0.863,0.782,0.796和0.931,0.825和0.641.预测模型具有良好的稳定性和预测能力.采用AutoDock4.2软件对药物与受体之间的结合方式进行了研究.运用这些信息能为进一步设计合成强效芳香化酶抑制剂,或筛选潜在的具有更强抑制活性的天然化合物提供帮助.
In this paper, molecular docking and 3D-HoVIAF was used to find the action mode of 36 letrozole derivatives and 34 anastrozole derivatives between aromatase. Here quantitative structure ac- tivity relationships （QSAR） model were built by stability and generalization ability of the model was dations. For the two groups of compounds, the multiple linear regression （MLR）. The estimation strictly analyzed by both internal and external vali- correlation coefficient of established MLS mode （Rcum）, leave-one-out cross-validation （Qcv）, predicted values versus experimental ones of external samples （Qox,） were 0. 863, 0. 782, 0. 796 and 0. 931, 0. 825 and 0. 641, respectively. The results indi- cated that QSAR models had both favorable estimation stability and good prediction capabilities. By u- sing AutoDock 4.2 software, the best action mode to brug between acceptor was found. These results revealed that both models have good predictive capability to guide the design and structural modifica- tion of homologic compounds.
出处
《分子科学学报》
CAS
CSCD
北大核心
2014年第5期403-409,共7页
Journal of Molecular Science
基金
国家自然科学基金资助项目(21275094)
陕西科技大学研究生创新基金资助项目
关键词
芳香化酶抑制剂
定量构效关系
三维全息原子场作用矢量
多元线性回归
分子对接
aromatase inhibitors
quantitative structure-activity relationship （QSAR）
3D holographicvector of atomic interraction filed （3D-HoVAIF）
multiple linear regression（MLR）
Molecular docking
