You are here:
Tox21 and ToxCast Chemical Landscapes: Laying the Foundation for 21st Century Toxicology
Citation:
RICHARD, A. M. Tox21 and ToxCast Chemical Landscapes: Laying the Foundation for 21st Century Toxicology. Presented at 5th Meeting on U.S.Govt. Chemical Databases and Open Chemistry, Frederick, MD, August 25 - 26, 2011.
Impact/Purpose:
Creating structure-activity-ready formats of actual test substance libraries linked to assay data poses several practical challenges (how to handle test duplicates, salts, stereo isomers, mixtures, etc). Cheminformatics, feature and property profiling, and a priori and interactive categorization of these libraries in relation to biological activity will serve as essential components of toxicity prediction
Description:
The U.S. Environmental Protection Agency’s ToxCast project and the related, multi-Agency Tox21 project are employing high-throughput technologies to screen hundreds to thousands of chemicals in hundreds of assays, probing a wide diversity of biological targets, pathways and mechanisms for use in predicting in vivo toxicity. The ToxCast chemical library consists of 960 unique chemicals and spans a diverse range of chemical structures and use categories. This library is fully incorporated into EPA’s approximately 4000 chemical contribution to the larger, more diverse Tox21 chemical library (totaling 10K). These chemical libraries represent central pillars of the ToxCast and Tox21 projects and are unprecedented in their scope, structural diversity, multiple use-scenarios (pesticides, industrial, food-use, drugs, etc.), and chemical feature characteristics in relation to toxicology. Chemical databases built to support these efforts consist of high quality DSSTox chemical structures and generic substance descriptions linked to curated test sample information (supplier, lot, batch, water content, analytical QC). Data files and supporting information (analytical QC summary data) will be publicly available in a variety of formats. Creating structure-activity-ready formats of actual test substance libraries linked to assay data poses several practical challenges (how to handle test duplicates, salts, stereo isomers, mixtures, etc). Cheminformatics, feature and property profiling, and a priori and interactive categorization of these libraries in relation to biological activity will serve as essential components of toxicity prediction strategies. [Abstract does not represent EPA policy. ]