电拓扑状态指数在黄酮衍生物作为HIV-1整合酶抑制剂的定量构效关系分析中的应用。
Application of the electrotopological state index to QSAR analysis of flavone derivatives as HIV-1 integrase inhibitors.
作者信息
Buolamwini J K, Raghavan K, Fesen M R, Pommier Y, Kohn K W, Weinstein J N
机构信息
Department of Medicinal Chemistry, School of Pharmacy, University of Mississippi 38677, USA.
出版信息
Pharm Res. 1996 Dec;13(12):1892-5. doi: 10.1023/a:1016005813432.
PURPOSE
A QSAR study based on electrotopological state (E-state) indices was conducted for a series of flavone HIV-1 integrase inhibitors to guide drug design.
METHODS
E-state indices formulated to encode electronic and topological information for each skeletal atom in a molecule (Kier and Hall Pharm. Res. 7:801-807 (1990)) were calculated using the Molconn-X program, and partial least squares (PLS) multivariate regression was used to derive QSAR models.
RESULTS
Predictive models with correlation coefficients (r2) of 0.98 (3 PLS components) and 0.99 (5 PLS components) and corresponding cross-validated correlation coefficients (c.v. r2) of 0.51 and 0.73, were obtained for inhibition of cleavage and integration, respectively, with one molecule omitted from the analysis.
CONCLUSIONS
E-state indices at C6, C3', C5', C5, and O4 were found to be more important for prediction of activity than those for any of the other 12 flavone skeletal atoms that are common to the molecules in the data set.
目的
基于电子拓扑状态（E-状态）指数对一系列黄酮类HIV-1整合酶抑制剂进行定量构效关系（QSAR）研究，以指导药物设计。
方法
使用Molconn-X程序计算为分子中每个骨架原子编码电子和拓扑信息的E-状态指数（Kier和Hall，《药物研究》7:801 - 807（1990）），并采用偏最小二乘（PLS）多元回归推导QSAR模型。
结果
对于切割抑制和整合抑制，分别获得了相关系数（r2）为0.98（3个PLS成分）和0.99（5个PLS成分）以及相应的交叉验证相关系数（c.v. r2）为0.51和0.73的预测模型，分析中省略了一个分子。
结论
发现C6、C3'、C5'、C5和O4处的E-状态指数对于活性预测比数据集中分子共有的其他12个黄酮骨架原子中的任何一个更重要。
Zotero 插件