摘要
将有机污染物分子中的非氢原子分为4类,将不同非氢原子及非氢原子之间的关系作为分子结构描述符,对49种有机污染物分子结构进行了参数化表达。采用逐步回归(SMR)与偏最小二乘回归(PLS)方法构建了有机污染物结构与生物活性之间的关系模型,模型的相关系数(R^2)为0.869,"留一法"交互检验相关系数(Q^2)为0.821,可见模型稳定性好、预测能力强,非偶然因素所致。
The organic molecule＇s non-hydrogen atoms were grouped into 4 categories. Different non-hydrogen atoms and their relationships were used as structural descriptors. 49 organic pollutants were parameterized. The step-wise regression （SMR） and partial least squares regression （PLS） methods were employed to build the relationship model between the structure of organic pollutants and biological activity. Results showed that the model correlation coefficient （i,2） was 0.869 and the “leave one out” cross validation correlation coefficient （Q2） was 0.821,indicating that the model was stable, predictive and not by chance.
出处
《环境污染与防治》
CAS
CSCD
北大核心
2017年第6期653-657,663,共6页
Environmental Pollution & Control
基金
四川省教育厅青年基金资助项目(No.13ZB0003)
四川省科技厅应用基础项目(No.2015JY0077)
关键词
有机污染物
生物活性
结构描述符
organic pollutants
biological activity
structural descriptor
