Abstract
A chemical safety database system named CAESAR (Computer-Aided Evaluation of Chemical Safety with QSAR) has been developed to assess the toxicity for diverse organic compounds. The system contains experimental toxicity data critically reviewed on selected chemical substances, and toxicity data predicted with QSAR models for numerous, diverse chemicals in commerce. As a first step, the carcinogenicity database has been constructed, where experimental carcinogenicity data on about 400 compounds were collected from various sources such as IARC, NTP and others, and their reliabilities were ranked into nine categories. To construct a predictive model of the carcinogenicity for diverse chemicals, 70 kinds of molecular descriptors were calculated from their 3D structures for these compounds, and the relationship between experimental carcinogenicity data and 25 parameters summarized from those descriptors with the principal component analysis was analyzed with an artificial neural network. The performance of the proposed model was assessed by applying the leave-one-out cross validation test. It was found that this method can predict the relative carcinogenicity of diverse chemicals with reasonable accuracy.
