苯酚及其衍生物对水生梨形四膜虫急性毒性的QSTR研究  

A QSTR study of the acute toxicity of phenol derivatives to aquatic Tetrahymena pyriformis

作　　者：张雅雄[1] 李琴[1] 杨彩蓉 

机构地区：[1]山西师范大学化学与材料科学学院,山西临汾041000

出　　处：《计算机与应用化学》2016年第2期209-212,共4页Computers and Applied Chemistry

基　　金：山西省留学回国人员项目(2014-045);山西省自然科学基金项目(2010011013-2);山西师大教改项目(SD2013JGXM-54)

摘　　要：选取了258个苯酚及其衍生物对水生梨形四膜虫的毒性数据,选择7个分子描述符作为建模的结构参数,开展了以QSTR方法建立苯酚及其衍生物毒性模型的研究。首先,运用稳健诊断方法(Robust Diagnostic Method)剔除奇异样本,然后采用球型排除算法(Sphere-exclusion Algorithms)合理划分样本,继而分别采用多元线性回归(Multiple Linear Regression,MLR)、偏最小二乘(Partial Least Squares,PLS)、BP(Error Back-Propagation,BP)神经网络3种方法进行定量构效关系研究,并对外部验证集采用共识建模方法(Consensus Modeling Method),从而提高了模型的预测能力。研究结果表明,所建模型均具有较好的预测能力和稳定性,且与MLR、PLS模型相比,BP神经网络模型性能略胜一筹,即非线性模型比线性模型性能优越。但是BP神经网络建立的模型不能直接给出直观的数学模型和公式,而MLR、PLS模型更为简单明了。A quantitative structure-toxicity relationship （QSTR） studies have made a lot of achievements in the field of environmental science. It is necessary to establish the QSTR models of the toxicity of phenol derivatives. 258 toxicity data of phenol derivatives to aquatic Tetrahymena pyriformis were selected form literature in this paper, and the QSTR models were established using 7 molecular descriptors. A robust diagnostic method was used to eliminate outliers, and the sphere exclusion algorithm was applied to divide the whole data set rationally. In the QSTR studies, multiple linear regression （MLR）, partial least squares （PLS） regression, and BP （Error Back- Propagation, BP） neural network were used as modeling tools, respectively. Then, for the external validation set, consensus modeling method was used to achieve more accurate predictions. According to the QSTR research results, the models show their high accuracy and good prediction abilities. Compared with MLR and PLS models, BP neural network models have their advantages. That is to say, nonlinear models are better than the linear ones for the QSTR research in this work. However, BP neural network models cannot give the mathematical formula directly. Moreover, MLR and PLS models are more simple.

关 键 词：苯酚及其衍生物 定量结构-毒性相关 毒性 共识建模法 

分 类 号：TQ015.9[化学工程] TP391.9[自动化与计算机技术—计算机应用技术]