堵锡华,
王超
徐州工程学院 化学化工学院, 徐州 221018

作者简介: 堵锡华(1963—),男,教授,研究方向为环境污染物构效学研究,E-mail:12dxh@sina.com.

基金项目: 国家自然科学基金项目(No.21472071,No.21473081)


中图分类号: X171.5

Quantitative Structure-Activity Model of Toxicity of Alcohol and Phenolic Pollutants to Rana temporaria Tadpoles and Tetrahymena pyriformis

Du Xihua,
Wang Chao
School of Chemistry and Chemical Engineering, Xuzhou University of Technology, Xuzhou 221018, China

CLC number: X171.5

-->

摘要
HTML全文
图(0)表(0)
参考文献(0)
相关文章
施引文献
资源附件(0)
访问统计

摘要:醇和酚类等有机化合物作为重要的工业原料,广泛应用于医药卫生、有机合成、食品工业等领域,但生产中排放于环境的这些物质,会对生物造成一定的毒性作用。为建立包含醇和酚类有机污染物对欧洲林蛙蝌蚪及梨形四膜虫毒性的定量结构-活性相关性模型,计算了227种有机污染物的分子连接性指数和分子形状指数,优化筛选了分子连接性指数的⁰X、¹X、²X、⁴X和⁵Xc、分子形状指数的K₁和K₂共7种参数,将这7种结构参数作为神经网络输入层变量,110种有机污染物对欧洲林蛙蝌蚪的毒性值作为输出层变量,采用7:8:1的网络结构方式,构建了令人满意的对欧洲林蛙蝌蚪毒性的神经网络预测模型,方程总相关系数r为0.988,毒性预测值与实验值之间的平均误差为0.14。为检验指数的普适性,同样用这7个结构参数与117种醇和酚类化合物对梨形四膜虫的毒性进行分析,所得神经网络模型的总相关系数达到0.997,对梨形四膜虫毒性的预测值与实验值之间的平均误差仅为0.065,结果表明,所建模型具有良好的预测有机污染物对林蛙蝌蚪及梨形四膜虫急性毒性的能力。
关键词: 有机污染物/
欧洲林蛙/
梨形四膜虫/
分子连接性指数/
分子形状指数/
神经网络

Abstract:Organic compounds, such as alcohols and phenols, were important industrial raw materials. They were widely used in pharmaceutical industry, organic synthesis industry, food industry, etc. However, emission of these organic compounds had toxic effects on organisms in the environment. In order to establish quantitative structureactivity relationship model of the toxicity of organic contaminants to Rana temporaria tadpoles and Tetrahymena pyriformis, the relationship between molecular structure of 227 kinds of organic contaminants and the toxicity to Rana temporaria tadpoles and Tetrahymena pyriformis was analyzed. Moreover, molecular connectivity indices and molecular shape indices of organic compounds were calculated. The molecular connectivity indices ⁰X, ¹X, ²X, ⁴X and ⁵X_c, and molecular shape indices K₁ and K₂, were selected. Then, the seven indices were used as input layer variables of neural network, the toxicity of 110 organic contaminants to Rana temporaria tadpoles was used as out-put layer variable and the 7:8:1 network structure was adopted to establish a satisfying neural network model. The total correlation coefficient r was 0.988. The mean error between the predicted value and experimental value was 0.14. In order to test universality, correlation between the structural parameters and the toxicity of 117 alcohol and phenolic compounds to Tetrahymena pyriformis was also analyzed by using the same method. The total correlation coefficient r was 0.997. The mean error between the predicted value and experimental value was 0.065. The results showed that the model had good predictive ability of the acute toxicity of organic contaminants to Rana temporaria tadpoles and Tetrahymena pyriformis.
Key words:organic contaminant/
Rana temporaria/
Tetrahymena pyriformis/
molecular connectivity index/
molecular shape index/
neural network.