摘要
根据定量结构-活性相关性(QSAR)原理,研究了38种取代芳烃对发光菌的急性毒性(Ce,50)与其分子结构之间的构效关系。应用遗传算法筛选出5个与Ce,50最为相关的描述符,并应用多元线性回归方法和支持向量机方法建立QSAR模型。两种模型的复相关系数、留一法交互验证系数分别为0.988、0.979和0.991、0.981,对外部预测样本的复相关系数和外部测试集交互验证系数分别为0.913、0.904和0.924、0.906,与相关文献比较,所建QSAR模型均具有更好的预测能力和稳健性。
This QSAR study relates to the structure of 38 sorts of substituted aromatic compounds,in which a set of five descriptors were chosen by using the variable selection method of genetic algorithm.The five descriptors were used to establish the QSAR of the acute toxicity of substituted aromatic compounds to photobacterium phosphoreum by multiple linear regression and support vector machine.The statistical results indicate that the multiple correlation coefficient and cross validation using leave-one-out were 0.988,0.979 and 0.991,0.981,respectively.To validate the predictive power of the resulting models,external validation multiple correlation coefficient and cross validation were 0.913,0.904 and 0.924,0.906,respectively.Compared with pertinent literature,the QSAR models have more favorable estimation stability and better prediction power.
出处
《常州大学学报（自然科学版）》
CAS
2012年第4期8-12,共5页
Journal of Changzhou University：Natural Science Edition
