多氯联苯正辛醇-水分配系数的定量构效关系  

QSPR study on n-octanol-water partition coefficient of polychlorinated biphenyls

作　　者：焦龙[1,2] 

机构地区：[1]西安石油大学化学化工学院,陕西西安710065 [2]西北大学化工学院,陕西西安710069

出　　处：《西北大学学报（自然科学版）》2011年第4期623-627,共5页Journal of Northwest University（Natural Science Edition）

基　　金：陕西省自然科学基金资助项目(2010JQ2003);陕西省教育厅科研计划基金资助项目(2010JK780);西北大学合成与天然功能分子化学教育部重点实验室开放基金资助项目(09014)

摘　　要：目的建立多氯联苯(polychlorinated biphenyls,PCBs)正辛醇-水分配系数(Kow)的定量结构性质关系(QSPR)模型。方法以5种量子化学参数和1种拓扑指数为结构描述符,运用偏最小二乘方法(PLS)建立QSPR模型,应用留一交叉验证及外部测试集验证两种方法对模型的预测能力进行评估。结果留一交叉验证中PLS模型预测12个化合物Kow的均方根相对误差(RMS%RE)为5.84,全部预测值与实验值间相关系数为0.884 1。外部测试集验证中该模型预测4个化合物Kow的RMS%RE为2.56,全部预测值与实验值间相关系数为0.820 3。结论该模型可用于PCBs的Kow值预测,PLS可用于建立多氯联苯Kow的QSPR模型。Aim To develope a quantitative structure property relationship （QSPR） model for predicting the n-octanol-water partition coefficient, Kow, of 16 PCBs. Methods Partial least squares （PLS） method was used to develop the calibration model for Ko~. The independent variables of PLS comprise five quantum structural descriptors and one topological index. Leave-one-out-cross validation and external test set validation were carried out to assess the predictive ability of the developed model. Results In the leave-one-out-cross validation, the RMS% RE of prediction is 5.84 and the correlation coefficient between the predicted and experimental log Kow is 0. 884 1. In the external test set validation, the RMS% RE of prediction is 2. 56 and the correlation coefficient between the predic- ted and experimental log Kow is 0. 820 3. Conclusion It is demonstrated that the developed model is applicable for predicting Kow of PCBs and PLS is a practicable method for developing QSPR model for Kow of PCBs.

关 键 词：定量结构性质关系 多氯联苯 正辛醇-水分配系数 偏最小二乘 

分 类 号：O6-39[理学—化学]