Scientists from around the world are invited to help develop computer models to predict toxicity. On Nov. 13, the Acute Toxicity Workgroup of the Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM) announced a global project, Predictive Models for Acute Oral Systemic Toxicity , which seeks new computer, or in silico, models.
The models should predict specific endpoints needed by regulatory agencies, such as whether a substance is either highly toxic or nontoxic. Computer models to predict toxicity are one of several approaches being explored by ICCVAM member agencies to reduce or replace animal use for acute systemic toxicity testing.
“We’re encouraging broad participation in this project, and we’re hoping to see groups using different approaches in their models,” said Nicole Kleinstreuer, Ph.D., deputy director of the NTP Interagency Center for the Evaluation of Alternative Toxicological Methods (NICEATM) and coordinator of the project. NICEATM supports ICCVAM, an interagency committee of the federal government that coordinates evaluation of new chemical safety testing approaches.
Regulatory agencies use acute systemic toxicity data to develop packaging requirements, product warning labels, personal protective equipment requirements, and environmental release guidelines.
Various approaches are needed
Participants in the project have access to a large body of toxicity data from rat studies collected by NICEATM and the U.S. Environmental Protection Agency (EPA) National Center for Computational Toxicology. On Nov. 17, NICEATM posted the data on the project website linked above. Modelers may use information such as chemical structure, chemical properties, and biological activity to predict these data.
Systemic toxicity is a complex process, and different regulatory agencies use data from these studies in various ways when evaluating chemical safety. Thus, no single modeling approach is likely to address all regulatory endpoints or predict toxicity for all classes of chemicals. The ICCVAM project will combine models that use different approaches to generate consensus predictions for a variety of endpoints.
Models will be presented at a workshop in April 2018 (see sidebar). A summary of the project models will be submitted for publication in the peer-reviewed literature, and the toxicity predictions generated by the models will be made available via the EPA Chemistry Dashboard .
(Catherine Sprankle is a communications specialist for ILS, the contractor supporting NICEATM.)