EPA Releases Improved Modeling Tool to Estimate Health Effects from Chemicals
Released December 5, 2024
Today, the U.S. Environmental Protection Agency (EPA) released of an updated and improved version of OncoLogic™, a modeling tool used to evaluate the potential of chemicals to cause cancer. EPA, in partnership with the Organisation for Economic Co-operation and Development (OECD), developed a more user-friendly version of the most widely used piece of this tool, greatly expanding the tool’s usability across the agency and the scientific community.
The tool has been updated to include subsystems that can evaluate fibers, metals, and polymers modules, and features:
- A streamlined interface that is much easier for non-experts to navigate;
- A standardized reporting format that allows users to quickly view and export results; and
- Increased transparency in the science behind the predictions provided by the model.
Also newly available is a plug-in that allows OncoLogic™ to be used in the OECD QSAR Toolbox versions 4.7 and above. The Toolbox contains measured data that OncoLogic™ uses for evaluating cancer concern, such as the results of genotoxicity tests. The Toolbox can also be used to evaluate chemicals in batch mode so multiple chemicals can be run at the same time.
OncoLogic™ is one of many publicly available assessment methods, databases, and predictive tools developed by EPA to estimate hazard to humans and the environment, particularly in the absence of test data. These tools and models support EPA staff analyses in implementing programs and regulations such as the Toxic Substances Control Act (TSCA) and help external users assess and manage chemical risks.
Download the OncoLogic™ model and supporting documentation.
Background
OncoLogic™ is a peer-reviewed predictive system that analyzes chemical structures to determine the likelihood that they might cause cancer. The model can evaluate more than 52 classes of organic chemicals, as well as fibers, metals, and polymers.
The OncoLogic™ model works by analyzing chemical and use information submitted by a user, then following a set of knowledge rules based on decades of research on how chemicals cause cancer in animals and humans including the known carcinogenicity of chemicals with similar chemical structures, information on mechanisms of action, short-term predictive tests, epidemiological studies, and expert judgment. Based on this analysis, the model constructs an estimation of the potential of chemicals to cause cancer, assigning a baseline concern level for the chemical.