卤代烷烃正辛醇/水分配系数的模拟  

Octanol-water partition coefficients simulation for halogenated hydrocarbons

作　　者：廖立敏[1] 李雪谊 邹宁[1] 雷光东[1] 

机构地区：[1]内江师范学院化学化工学院,四川内江641199

出　　处：《计算机与应用化学》2014年第5期619-622,共4页Computers and Applied Chemistry

基　　金：2013年四川省教育厅青年基金项目(13ZB0003)

摘　　要：采用分子电性相互作用矢量(MEIV)对36个卤代烷烃化合物进行了结构表征,分别采用神经网络技术和多元线性回归方法构建了该类化合物正辛醇/水分配系数(IgKow)的预测模型。两模型的复相关系数(R)分别为0.900和0.945,标准偏差(SD)分别为0.233和0.188。结果表明所建模型具有较好的稳健性和预测能力,同时也证明了分子电性相互作用矢量在卤代烷烃类化合物QSPR研究中的适用性。In this paper, the structures of 36 halogenated hydrocarbons were characterized by the molecular electronegativity interaction vector （MEIV）. Neural network and multiple linear regression methods were employed to construct prediction models of octanol / water partition coefficient （lgKow） of the compounds. The multiple correlation coefficients （R） of the Neural network model and MLR model were 0.900 and 0.945, and the standard deviations （SD） of the two models were 0.070 and 0.064, respectively. The results showed good robustness and predictive ability of the proposed models, and satisfactory applicability of the molecular electronegativity interaction vector （MEIV） in the QSPR study of halogenated hydrocarbons.

关 键 词：卤代烷烃 醇 水分配系数 结构描述子 定量结构性质相关 

分 类 号：O625.21[理学—有机化学]