文献类型：期刊文献

英文题名：In silico prediction of Tetrahymena pyriformis toxicity for diverse industrial chemicals with substructure pattern recognition and machine learning methods

作者：Cheng, Feixiong[1];Shen, Jie[1];Yu, Yue[1];Li, Weihua[1];Liu, Guixia[1];Lee, Philip W.[1,2];Tang, Yun[1]

机构：[1]E China Univ Sci & Technol, Sch Pharm, Dept Pharmaceut Sci, Shanghai 200237, Peoples R China;[2]Kyoto Univ, Grad Sch Agr, Sakyo Ku, Kyoto 6068502, Japan

年份：2011

卷号：82

期号：11

起止页码：1636

外文期刊名：CHEMOSPHERE

收录：;WOS:【SCI-EXPANDED(收录号:WOS:000288587900017)】；

基金：This work was supported by the Program for New Century Excellent Talents in University (Grant No. NCET-08-0774), the 863 High-Tech Project (Grant No. 2006AA020404), the 111 Project (Grant No. B07023), and the National S&T Major Project of China (Grant No. 2009ZX09501-001).

语种：英文

外文关键词：Tetrahymena pyriformis toxicity; Quantitative structure-toxicity relationship; Substructure pattern recognition; Support vector machine; Machine learning; Information gain

摘要：There is an increasing need for the rapid safety assessment of chemicals by both industries and regulatory agencies throughout the world. In silico techniques are practical alternatives in the environmental hazard assessment. It is especially true to address the persistence, bioaccumulative and toxicity potentials of organic chemicals. Tetrahymena pyriformis toxicity is often used as a toxic endpoint. In this study, 1571 diverse unique chemicals were collected from the literature and composed of the largest diverse data set for T. pyriformis toxicity. Classification predictive models of T. pyriformis toxicity were developed by substructure pattern recognition and different machine learning methods, including support vector machine (SVM), C4.5 decision tree, k-nearest neighbors and random forest. The results of a 5-fold cross-validation showed that the SVM method performed better than other algorithms. The overall predictive accuracies of the SVM classification model with radial basis functions kernel was 92.2% for the 5-fold cross-validation and 92.6% for the external validation set, respectively. Furthermore, several representative substructure patterns for characterizing T. pyriformis toxicity were also identified via the information gain analysis methods. (C) 2010 Elsevier Ltd. All rights reserved.