High throughput in vitro hazard assessment is an integral part of chemical prioritization in the 21st century risk assessment framework that calls for the design of alternative methods to animal testing. Increased confidence in in vitro assay interpretation requires that assays are suitable for a compound’s properties. Dosimetry in in vitro systems can be impacted substantially by physicochemical properties of the test agent, leading to systematic bias in assays. Chemical space analysis interrogates chemical properties across multiple determinants of behavior and informs chemical compatibility with test systems. The majority of current high throughput in vitro models were originally designed for pharmaceutical compound screening. These methods can be applied to a set of environmental compounds that share similar chemical space in order to fill data gaps, but chemicals with widely disparate structures may be outside the domain of applicability. We identified a set of active pharmaceutical ingredients among ~45,000 curated chemical structures from the Collaborative Estrogen Receptor Predictor Project (CERAPP), which includes compounds from Tox21, ToxCast, and the Endocrine Disruptor Screening Project (EDSP). This set of active pharmaceutical ingredients were then compared with a subset of CERAPP compounds identified as being pharmaceutical, near-field, or far-field environmental chemicals in the USEPA Aggregated Computational Toxicology Resources (ACToR) database. CERAPP compounds were ranked by their chemical similarity to the top pharmaceutical actives, using physicochemical properties and ToxPrint chemotypes as descriptors. The resulting Jaccard-Tanimoto indices show that environmental compounds that were classified as either near-field or far-field compounds per ACToR, were more similar to each other than they were to those classified as top pharmaceutical ingredients. A set of 15 compounds were selected for further in vitro studies that take into account the overlap between the environmental and pharmaceutical domains. This work does not reflect U.S. EPA policy.

Applying chemical space analysis to select compounds for targeted in vitro methods

255th National Meeting (2018)

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