Enhancing toxicological interpretation and reporting of

Introduction: A wide variety of in vitro high-throughput screening (HTS) assay data are publicly available and have the potential to serve as the foundation for the development of computational toxicology approaches. However, the sources of these data vary in the format and metadata they provide and the clarity of their relevance to toxicological endpoints. Thus, applying them to decisions about chemical effects remains a challenge. To overcome this challenge and help democratize HTS data, we curated and annotated HTS data specifically for toxicological interpretation within a regulatory endpoint context. Methods: Focusing on data from the publicly available ToxCast/Tox21 HTS program, a curation workflow was developed to identify the most reliable data by integrating metadata such as chemical quality control data, concentration-response curve-fitting robustness, assay-specific limitations, and potential technological interference. To this curated HTS (cHTS) dataset, we then applied structured vocabularies and expert review to develop molecular/cellular process annotations mapping HTS assays to toxicologically relevant regulatory endpoints. These data were also mapped to the Organization for Economic Co-operation and Development's Harmonised Template 201 (OHT 201) reporting format to standardize reporting, rendering outputs amenable to regulatory submission. Results: The curated and annotated cHTS dataset supports data interoperability and allows these data to be searched, grouped, and visualized by regulatory endpoints (e.g., endocrine disruption, carcinogenicity, developmental toxicity, and systemic effects). The robust cHTS data and annotations can be retrieved from the publicly accessible National Institute of Health (NIH) Integrated Chemical Environment (NICE, https://ice.ntp.niehs.nih.gov/). Access through NICE allows these data to be optimally leveraged for a variety of uses, from identifying data gaps and providing mechanistic insights to confidently and reproducibly retrieving data, thus supporting their integration to build weight-of-evidence assessments for regulatory-relevant toxicological endpoints. Discussion: Developing an environment in which HTS data that are structured, accessible, and readily interpretable in a toxicological context, ensures that users can apply these evolving resources. Working toward consistent data summaries by leveraging established reporting formats helps bolster regulatory interpretation and adoption. Overall, the work summarized herein demonstrates how fit-for-purpose standardized curation, annotations, ontologies, ease of accessibility, and structured reporting can support data interoperability and the application of complex HTS data for regulatory application.