Data Integration and Visualization for Transparent Communication in Read-Across: a Glycol Ether Case Study

Abstract

Integration and communication of diverse information into human health assessments is often challenging given the wide variety of data types and sources under consideration. In this study, several new approaches are utilized for compiling and visualizing chemical structure and toxicological assay-derived information in support of both category and analogue read-across. A case study of propylene-based (P-series) glycol ethers was chosen, representing high production volume chemicals with a large database that includes physicochemical properties and various measures of mammalian toxicity. As a comparator, ethylene-based (E-series) glycol ethers were used to examine the potential for category separation and broader structural groupings. Two approaches for integrating and visualizing these complex data were used, Toxicological Priority Index (ToxPi) and Chemical-Biological Read-Across (CBRA), which allow for the unique incorporation of various types of data, differential weighting, relative ranking, and step retention to maximize transparency. The raw toxicity study data were transformed, scaled, and rendered to compose visualizations with alternate category grouping scenarios. We found overall that the glycol ethers group together within their structure-based category, as well as within series. Both ToxPi and CBRA facilitate effective communication of complex data and enable groupings and selection of most suitable analogues for read-across.