酚类化合物对梨形四膜虫毒性的构效关系解析  

Studies on QSAR for the toxicity of phenols against tetrahymena pyriformis

作　　者：顾云兰[1,2] 张淦杰 

机构地区：[1]江苏省滩涂生物资源与环境保护重点建设实验室,江苏盐城224051 [2]盐城师范学院化学与环境工程学院,江苏盐城224051

出　　处：《化学研究与应用》2016年第6期810-817,共8页Chemical Research and Application

基　　金：江苏高校品牌专业建设工程资助项目(PPZY2015B113);江苏省滩涂生物资源与环境保护重点建设实验室开放基金资助课题(JLCBE12009)

摘　　要：应用密度泛函理论(DFT)在杂化泛函B3LYP和基组6-311G**水平上对81个标题化合物进行结构优化,计算获得化合物的量化和理化参数,采用最佳变量子集回归构建化合物对梨形四膜虫的毒性与描述符之间的定量结构-活性关系(QSAR)模型。所建最佳四元QSAR模型的复相关系数R^2为0.951,去一法(LOO)交互检验复相关系数R^2_(cv)为0.938,用27个验证集样本进行外部预测,所得外部验证集交互检验系数Q_(ext)~2为0.876,模型具有良好的统计学意义及合理、可信的预报能力,影响毒性的主要因素有正辛醇-水分配系数(log P)、分子最低空轨道能(E_(LUMO))、苯环上所有碳原子电荷之和(Q)和4号位取代基电荷(Q_(R4)),研究结果为理解该类化合物的毒性作用机理和设计合成低毒性的新化合物提供理论参考。The geometrical structures of phenols were optimized and the corresponding quantum-chemical parameters and physicalchemical parameters were calculated using DFT/B3 LYP procedure at 6-311G＊＊level. A reasonable QSAR equation for predicting the toxicity of phenols to tetrahymena pyriformis was achieved with a quite high complex correlation coefficient（ R^2= 0. 951） by leaps-and-bounds regression.The performance of the model was tested through cross-validation using leave-one-out procedure（ LOO）and external validation with R^2_（cv）and Q^2_（ext）values of 0. 938,0. 876,respectively. The results show that not only the QSAR model has good stability and predictability,but also the octanol-water partition coefficient（ log P）,the energy of the lowest unoccupied orbit（ E_（LUMO））,the sum of the charge of carbon atoms on the benzene ring（ Q） and the charge of 4 substituent（ Q_（R4）） are the main factors affecting the toxicity of phenols.The developed QSAR model can be used to predict the toxicity of phenolic compounds to tetrahymena pyriformis.The results present important information for the toxicitical mechanism and will be useful for theoretical designing new title compounds with low toxicty.

关 键 词：酚类化合物 梨形四膜虫 密度泛函理论 最佳变量子集回归 定量构效关系 

分 类 号：O641[理学—物理化学]