Lu Chunhui, Wang Yang, Yin Chunsheng, Guo Weimin, Hu Xiaofang
School of Environmental Science and Engineering, Shanghai Jiao Tong University, 800 Dongchuan Road, Min Hang, Shanghai 200240, PR China.
Chemosphere. 2006 May;63(8):1384-91. doi: 10.1016/j.chemosphere.2005.09.052. Epub 2005 Nov 22.
Quantitative structure-property relationship (QSPR) models of soil sorption coefficients for 32 persistent organic pollutants were constructed using our recently introduced Lu index and novel distance-based atom-type DAI topological indices. Using multiple linear regression technique, a 6-variable model was obtained with the correlation coefficient of estimations (R) being 0.95, and the standard error of estimations (s) being 0.23, and the correlation coefficient (R(cv)) and the standard error (s(cv)) in the leave-4-out cross-validation procedure are 0.90 and 0.31, respectively. The results in this study indicate that soil sorption coefficients of POPs are dominated by molecular size while some DAI indices have smaller influence.
利用我们最近引入的鲁指数和新型基于距离的原子类型DAI拓扑指数，构建了32种持久性有机污染物土壤吸附系数的定量结构-性质关系（QSPR）模型。采用多元线性回归技术，得到了一个六变量模型，估计相关系数（R）为0.95，估计标准误差（s）为0.23，在留一法交叉验证过程中的相关系数（R（cv））和标准误差（s（cv））分别为0.90和0.31。本研究结果表明，持久性有机污染物的土壤吸附系数主要受分子大小的影响，而一些DAI指数的影响较小。
