Contained within the 28 million citations in PubMed is an abundance of information about the activity of chemicals in biological systems. At the EPA’s National Center for Computational Toxicology we have implemented literature informatics approaches that make effective use of the literature in a variety of applications important to environmental science. For instance, we use text-mining methods to identify chemicals that may cause adverse effects like neurotoxicity or embryonic vascular disruption. Then we can take the resulting list of chemicals and, using the PubMed Abstract Sifter tool, we can explore the linkages and connections to key events in defined adverse outcome pathways (AOPs).  Quantitative alterations in embryonic vascular disruption and dysmorphogenesis can be expanded into an AOP for developmental toxicity utilizing information extracted by the Abstract Sifter. The Abstract Sifter technology has been implemented in two forms.  One version integrates the richness of PubMed with the flexible data-manipulation capabilities of Microsoft Excel. A web-based implementation of the technology is available through the EPA CompTox Chemicals Dashboard [https://www.epa.gov/chemical-research/comptox-chemicals-dashboard]. These literature informatics applications help chemists and toxicologists take advantage of the immense knowledge recorded in the chemical literature for predictive toxicology. <i>This abstract does not necessarily represent U.S. EPA policy.</i><br/>

Literature-based cheminformatics for research in chemical toxicity

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.