文献类型：期刊文献

英文题名：In Silico Estimation of Chemical Carcinogenicity with Binary and Ternary Classification Methods

作者：Li, Xiao[1,3];Du, Zheng[1];Wang, Jie[1];Wu, Zengrui[1];Li, Weihua[1];Liu, Guixia[1];Shen, Xu[3];Tang, Yun[1,2]

机构：[1]E China Univ Sci & Technol, Sch Pharm, Shanghai Key Lab New Drug Design, Shanghai 200237, Peoples R China;[2]Shanghai Tobacco Grp Co Ltd, Ctr Tech, Key Lab Cigarette Smoke, Shanghai 200082, Peoples R China;[3]Chinese Acad Sci, Shanghai Inst Mat Med, Shanghai 201203, Peoples R China

年份：2015

卷号：34

期号：4

起止页码：228

外文期刊名：MOLECULAR INFORMATICS

收录：;WOS:【SCI-EXPANDED(收录号:WOS:000352620100004)】；

基金：This work was supported by the National Natural Science Foundation of China (Grant 81373329), the 863 Project (Grant 2012AA020308), Shanghai Tobacco Group Co. Ltd. Research Fund (Grant K2013-1-044P), and the Fundamental Research Funds for the Central Universities (Grant WY1113007).

语种：英文

外文关键词：Binary classification; Ternary classification; Chemical carcinogenicity; Machine learning methods; Tobacco smoke

摘要：Carcinogenicity is one of the most concerned properties of chemicals to human health, thus it is important to identify chemical carcinogenicity as early as possible. In this study, 829 diverse compounds with rat carcinogenicity were collected from Carcinogenic Potency Database (CPDB). Using six types of fingerprints to represent the molecules, 30 binary and ternary classification models were generated to predict chemical carcinogenicity by five machine learning methods. The models were evaluated by an external validation set containing 87 chemicals from ISSCAN database. The best binary model was developed by MACCS keys and kNN algorithm with predictive accuracy at 83.91 %, while the best ternary model was also generated by MACCS keys and kNN algorithm with overall accuracy at 80.46 %. Furthermore, the best binary and ternary classification models were used to estimate carcinogenicity of tobacco smoke components containing 2251 compounds. 981 ones were predicted as carcinogens by binary classification model, while 110 compounds were predicted as strong carcinogens and 807 ones as weak carcinogens by ternary classification model. The results indicated that our models would be helpful for prediction of chemical carcinogenicity.