ONTOX Insights #5

The first ONTOX Hackathon took place on 21–23 April 2024 in Utrecht, The Netherlands, bringing together experts from science, industry, academia, and animal welfare to accelerate innovation in human safety through AI and reduce reliance on animal testing. Eight teams worked on challenges related to chemical risk assessment, AI integration, and the shift to human-relevant methods. The event concluded with a pitching contest, and the most promising ideas will be presented to regulatory authorities, industry, academia and non-governmental organizations at the next ONTOX Stakeholder Network meeting and taken up by the ONTOX project in order to tackle the above-mentioned challenges further. This report summarizes key lessons from the hackathon and presents the outcomes of the ONTOX Hackathon that resulted in new innovative solutions based on New Approach Methodologies (NAMs).

The Society for the Advancement of AOPs Knowledgebase Interest Group (SKIG) brings together over 40 international experts to advance the Adverse Outcome Pathways (AOP) framework. Through regular online meetings, SKIG explores key scientific and technical topics—especially AOP-Wiki-related issues—ranging from ontology harmonization to AI tools and omics integration. This report summarizes discussions and practical updates from meetings held in 2023–2024 and highlights SKIG’s continued contributions to supporting EU goals for animal-free chemical safety assessment.

Current in vivo tests for developmental neurotoxicity (DNT) are low-throughput, resource-heavy, and limited in human relevance. This study addresses the challenges by providing a comprehensive transcriptomic comparison of two human-induced pluripotent stem cells (hiPSC)-derived models, a 2D adherent and a 3D neurosphere, with differentiation period up to 21 days. Both models show maturation into post-mitotic neurons, with the 3D model favoring development of the GABAergic neurons, and the 2D models favoring glutamatergic neuron. Their profiles align with human fetal brain data, confirming physiological relevance. These findings support the complementary use of both models for DNT testing, highlighting the 2D model’s strength in synaptogenesis and the 3D model’s suitability for neural network formation. This research marks a significant advancement in developing human-relevant, high-throughput DNT assays for regulatory purposes.

The ONTOX project, funded by the EU Horizon 2020 and coordinated by Vrije Universiteit Brussel, Belgium, aims to provide a functional and sustainable solution for advancing human risk assessment of chemicals without the use of animals. This project is in alignment with AI and ontology-driven tools. As a proof-of-principle, ONTOX will conduct a probabilistic risk assessment (PRA) for perfluorooctanoic acid (PFOA), combining exposure and hazard data from literature, in vitro studies, and custom-made methods using R. The case study will apply the full ONTOX toolbox, including physiological based kinetic (PBK) models, quantitative in vivo in vitro extrapolation (QIVIVE), physiological maps (PMs), boolean models, and a large transformer-based AI model, to demonstrate a human-relevant, next-generation approach to risk assessment. This is a protocol for the case study on PFOA, which will provide a proof-of-principle of PRA using in vitro studies.

Follow us on social media!