摘要
以分子电性距离矢量(MEDV)描述子有效表征43个取代芳烃类化合物的分子结构,应用基于预测变量的选择与模型化(VMSP)方法,建立化合物的封闭体系绿藻(Pseudokirchneriella subcapitata)毒性(48 hEC50)与其分子结构之间的定量相关(QSTR)模型.应用多元线性回归方法建立的QSTR模型具有较高的相关系数(r=0.8911)及LOO(Leave-One-Out)检验相关系数(q=0.8102),表明该模型具有良好的估计能力与稳定性.应用28个化合物的训练集样本构建QSTR模型预测外部检验集,结果表明,训练集模型也具有良好的预测能力.
Molecular Electronegativity Distance Vector （MEDV） for atom types was used to describe the structures of 43 substituted aromatic hydrocarbon chemicals, and a quantitative structure-toxicity relationship （QSTR） between the MEDV descriptors and the algal （Pseudokirchneriella subcapitata） toxicity （48 h EC50 ） in a closed system was developed using variable selection and modeling based on prediction （VSMP）. A QSTR model with a high calibrated correlation coefficient of r = 0.891 1 and validated correlation coefficient of q = 0. 810 2 were built by multiple linear regression, which showed a good estimation ability and stability of the model. The good predictive power for the external samples was validated by the model built from the training set with 28 substituted aromatic hydrocarbon chemicals.
出处
《环境科学研究》
EI
CAS
CSCD
北大核心
2009年第7期823-827,共5页
Research of Environmental Sciences
基金
国家水体污染控制与治理科技重大专项(2008ZX08526-003)
关键词
分子电性距离矢量(MEDV)
封闭体系藻毒性
取代芳烃类化合物
定量结构-毒性相关
Molecular Electronegativity Distance Vector （MEDV）
algal toxicity in the close system
substituted aromatic hydrocarbon
quantitative structure-toxicity relationship （QSTR）
