电性拓扑指数预测Cedarmycins衍生物的杀菌活性

Prediction of Fungicidal Activity of Cedarmycins Derivatives to Phytophthora capsici Using Electrotopological State Indices of Atom Types

基于化学拓扑理论,计算了22种Cedarmycins衍生物的原子类型的电性拓扑状态指数(Em)。运用多元统计回归方法,建立上述有机物对辣椒疫霉FA与Em的QSAR模型。经逐步回归,建立了它们的三元数学模型,传统的判定系数(R2)为0.931,逐一剔除法的交互验证系数(R2cv)为0.842。该模型经R2cv、VIF检验,具有良好的稳健性及预测能力。根据进入模型的三个变量(E1、E5、E14)可知,影响衍生物杀菌活性的主要结构因素是分子中:甲基、醚氧键及在芳环内的=CH-等基团。

The molecular electrotopological state indices of atom types （Em ） of 22 cedarmycins derivatives were calculated by chemical topological method. The quantitative structure -activity relationships （QSAR） was established by using stepwise regression analysis for the fungicidal activity （ FA ） of these compounds to Phytophthora capsici along with the Em. The traditional correlation coefficient （R^2 ） and the cross - validation correlation coefficient （ R^2 cv ） of leave - one - out （LOO） were 0. 931 and 0. 842, respectively. The QSAR models had both favorable estimation stability and good prediction capability by R^2cv, VIF tests. From the three parameters of E1, E5. and E14 of the model, it showed that the molecular structure characteristics of two - dimensional, such as the structural fragment - CH3 , = CH - （ in benzene ring）, - O - , were the decisive factors affecting the fungicidal activity.