分子电性距离矢量预测多氯萘的正辛醇/空气分配系数

PREDICTING THE OCTANOL-AIR PARTITING COEFFICIENT OF PCNs USING MOLECULAR ELECTRONGATIVITYDISTANCE VECTOR

应用分子电性距离矢量(MEDV)对多氯萘(PCNs)的76种同系物进行结构表征,通过多元线性回归方法建立了PCNs的正辛醇/空气分配系数(lgKOA)与MEDV之间的定量结构-性质关系(QSPR),该模型具有显著的相关性(n=24,R=0.997,SD=0.088,F=417.546).采用逐步回归的方法(SMR)从原模型参数中选取2个参数建立了新模型,其模型相关系数R=0.996,继以留一法进行交互检验,相关系数RCV=0.995,说明此定量构效相关模型具有很好的稳定性和预测能力.运用模型对未有试验值的52种PCNs的lgKOA进行了预测.

Molecular electrongativity-distance vector（MEDV） was used to describe the chemical structures of 76 polychlorinated naphthalenes（PCNs）.A reasonable quantitative relationship model between the lgKOA of PCNs and the molecular electrongativity-distance vector（MEDV） was achieved by multiple linear regression（MLR）.The results of significance test were quite satisfactory（n=24,R=0.997,SD=0.088,F=417.546）.A more predictive model with high correlation coefficient（R=0.996） was constructed by selecting two parameters form all the elements in the MEDV vectors of the former model through stepwise multiple regression（SMR）.The performance of the two-parameter model was tested through cross-validation by the leave-one-out procedure（LOO） and satisfactory results were obtained（RCV=0.995）.Then octanol-air partition coefficients（lgKOA） of 52 PCNs with unknown experimental values were predicited by the models.