Development of Knowledge Rules for Cancer Expert System for Prediction of Carcinogenic Potential of Chemicals: USEPA Approach

Predictive Toxicology of Chemicals: Experiences and Impact of AI Tools

Volume

Predictive Toxicology of Chemicals: Experiences and Impact of AI Tools

Contents

Since the passage of the Toxic Substances Control Act (TSCA), which requires the U.S.Environmental Protection Agency (EPA) to assess potential toxic effects of new and existing chemicals, structureactivity relationships (SAR) analysis has been extensively and effectively used in the identification of potential health hazards of new (Premanufacturing Notification) industrial chemicals (Arcos, 1983; Auer and Gould, 1987). To evaluate the potential health hazards of the large number of existing chemicals for which adequate test data are not available, SAR analysis has been given an increasing role as a basis for recommending additional testing, for designing strategic research plans in TSCA programs, and for setting priorities of testing environmental pollutants such as disinfection byproducts in drinking water (Woo et al., 1999). To carry out pollution prevention initiatives, SAR has also been used to search for safer chemical substitutes and in molecular design of new chemicals (DeVito and Garrett, 1996). Most recently, SAR was used for the first time to provide support for the regulatory decision to list 2,4,6-tribromophenol as a hazardous waste (EPA, 1998).