P03-15 From Pathways to Predictions: An Ontology-Driven Data Integration Framework for Integrating AOP and PBPK in Chemical Risk Assessment

Chemical risk assessment demands a harmonized integration of qualitative and quantitative methodologies to address challenges in exposure modeling, hazard prediction, and regulatory decision-making. Ontologies, widely recognized for their role in structured data mining and knowledge representation, offer a powerful solution for formalizing biological knowledge and enabling machine-readable workflows. This work undertaken within the EU Partnership for Risk Assessment (PARC) presents an ontology-driven framework that bridges Adverse Outcome Pathways (AOPs) and Physiologically Based Pharmacokinetic (PBPK) models to advance chemical risk assessment through consistent data integration and automation. Physiologically Based Pharmacokinetic Ontology (PBPKO), comprising approximately 700 terms, has been developed to annotate PBPK models in SBML format. Submitted to the OBO Foundry and undergoing revision, PBPKO standardizes toxicokinetic terminology, facilitating interoperability across regulatory applications. Case studies on selected PBPK models demonstrate its utility for annotation of model with tools like PBK- inspector and harmonized workflow. This ontology enhances translational modeling by supporting quantitative in vitro - in vivo extrapolation (QIVIVE) and probabilistic exposure predictions. The development of an AOP ontology led by the PARC partners in collaboration with major AOPs stakeholders and regulatory agencies, aims to formalize AOP knowledge representation at a granular level, maintaining a balance between information depth and abstraction. The foundational terms and logical structure will be established to support future quantitative applications of AOP. Harmonizing qualitative and quantitative methods used in chemical risk assessment enables consistent reporting and facilitates knowledge sharing among regulatory bodies. This ontology-driven harmonization represents a significant step toward unifying fragmented data streams in chemical risk assessment. It enables systematic evidence evaluation, reduces reliance on animal testing through mechanistic insights, and supports regulatory decisionmaking by providing a cohesive framework for integrating exposure, toxicokinetics, and hazard information. Ongoing case studies highlight the potential of this approach to address emerging challenges in chemical safety assessment while fostering collaboration among stakeholders in academia, industry, and regulatory agencies.