芳香族化合物生物降解性的QSBR研究

分别采用线性基团贡献法和人工神经网络法对芳香族化合物的生物降解最大去除率QTOD进行QSBR研究。得到不同基团对生物降解性的贡献顺序为 :C6H5>COOH >OH >CO >CH3 >C1 >NH2>NO2 。线性基团贡献法对于训练组和测试组的预测正确率分别为 86%和 80 % ,总的预测正确率达85 % ;而人工神经网络法的预测正确率分别为 94%、80 %和 92 %。结果表明 ,线性基团贡献法和神经网络法的预测效果均很好 ,而神经网络法的预测更精确。

QSBR Study on Biodegradability of Aromatic Compounds

The quantitative structure biodegradability relationship studies were performed with the maximum specific removal rates ( Q TOD ) of aromatic compounds by the linear group contribution method and artificial neural network approach, respectively. The order of contribution of various group to biodegradation is C 6H 5>COOH>OH>CO>CH 3>Cl>NH 2>NO 2.The accuracy of prediction of the linear group contribution method is 86% for the training set, 80% for the test set, and 85% for all compounds. When by neural network approach, it is 94%, 80%, and 92%, respectively. It has been shown that both the linear group contribution method and the neural network method are able to fit very well whether for the training set or for the test set. However, the neural network method can provide a superior fit to biodegradability and give a lower prediction error than the linear regression method.