MCP Server Proto-OKN

BioBricks ToxCast is an open knowledge graph for computational toxicology researchers, regulatory scientists, and environmental health professionals seeking standardized access to EPA ToxCast high-throughput screening data. The graph contains 3.34 million bioassay screening results linking 9,542 chemical entities to 2,205 distinct assay endpoints, enabling systematic exploration of chemical-bioactivity relationships across diverse toxicological targets. Each chemical is identified by both EPA DSSTox Substance Identifiers (DTXSID) and InChIKeys, supporting cross-reference to external chemical databases. The knowledge graph employs standardized vocabularies including the BioAssay Ontology (BAO), Chemical Information Ontology (CHEMINF), and Biolink, ensuring semantic interoperability with other toxicology and biomedical knowledge graphs. Data is derived from EPA's public domain ToxCast database and released under CC0-1.0. Users can query the graph via SPARQL at the FRINK infrastructure endpoint to integrate ToxCast screening data with complementary datasets for compound prioritization, adverse outcome pathway analysis, and mechanism-based risk assessment.