昆虫酚氧化酶抑制剂的活性预测模型  

Prediction Model of the Activity of Insect Henoloxidase Inhibitors

作　　者：陈艳[1] 王思华[1] 李靖[1] 

机构地区：[1]徐州工程学院化学化工学院,江苏徐州221111

出　　处：《武汉大学学报（理学版）》2016年第3期293-298,共6页Journal of Wuhan University:Natural Science Edition

基　　金：国家自然科学基金资助项目(21272095)

摘　　要：采用电性拓扑状态指数(En)表征昆虫酚氧化酶(PO)抑制剂的分子结构,通过最佳变量子集回归的方法建立了57种PO抑制剂抑制活性(pIC50)的多元线性回归模型,非交叉相关系数和交叉相关系数分别为0.920和0.908,经Jackknife和变异膨胀因子(VIF)检验具有良好的稳定性和预测能力.该模型显示影响PO抑制剂抑制活性的主要因素是—OH,—O—和C=O等分子结构片段.以模型中的3个参数E13,E14,E16为人工神经网络输入层,设定3∶6∶1的网络结构构建人工神经网络的BP算法模型,相关系数达到0.988.结果表明,与多元线性回归模型相比,BP人工神经网络模型的相关性和预测能力均有较大的提高.Electrotopological state index （En）was used to describe the molecular structure of 57 insect henoloxi- dase inhibitors in this paper. The multiple liner regression（MLR） model was constructed by leaps-and-bounds regres- sion （LBR）, and the traditional correlation coefficient and the cross-validation correlation coefficient are 0. 920 and 0. 908, respectively. The model exhibited excellent stability and predictability evaluated by Jackknife and variance infla- tion factors （VIF） method. The model shows that the dominant influencing factors of inhibited activity are the molecu- lar structure fragments： -OH, -O-, and C = O in PO inhibitors. The three structural parameters E13, El4 , E16 were used as the input neurons of artificial neural network,and a 3 ： 6 ： 1 network architecture is employed. A satis- fied model can be constructed with the back-propagation algorithm, the correlation coefficient R2 was 0. 988. The re- suits show that the back-propagation algorithm of artificial neural network model have higher correlation and predicta- bility than multiple liner regression model.

关 键 词：电性拓扑状态指数 酚氧化酶抑制剂 人工神经网络 定量结构-活性相关 

分 类 号：O641.12[理学—物理化学]