Chairs: Jian Jiang & Jinhee Choi
Traditional toxicity testing remains heavily reliant on animal studies, underscoring the urgent need for innovative, human-relevant and ethically responsible alternatives. The rapid advancement of artificial intelligence (AI) is transforming toxicology by enhancing the predictive power, efficiency and mechanistic relevance of in vitro toxicity screening. AI-driven approaches, including machine learning, deep learning and explainable AI, provide powerful tools to improve in vitro-to-in vivo extrapolation, identify complex toxicity patterns and strengthen risk assessments. This session highlights how AI accelerates the development of New Approach Methodologies (NAMs), supporting the 3Rs (Replacement, Reduction and Refinement) and advancing regulatory acceptance. The invited experts will address key challenges in AI-driven toxicology, including the development of next-generation NAMs through advanced computational approaches, overcoming hurdles in data integration, standardization and cross-species translation for reliable toxicity predictions, bridging in vitro findings with human-relevant mechanisms to enhance risk assessment, and improving transparency, reproducibility and regulatory acceptance of AI-powered methodologies. The first presentation will explore large language models as interfacing layers to connect diverse data sources, aiding risk assessors in regulatory decision-making (ONTOX). The second will discuss AI-driven virtual control groups (VCGs) as an alternative to conventional control groups (CCGs) in preclinical animal studies, potentially reducing animal use by up to 25% (VICT3R). The third will highlight AI-based in silico methods for predicting toxicity and ADME properties, supporting drug and chemical safety assessments (RISK-HUNT3R). The fourth will explore AI-driven network-based methods for translating model organism toxicity data into human-relevant insights, enhancing cross-species translation and mechanistic interpretation (PrecisionTox). This session fosters dialogue among researchers, industry leaders and regulatory authorities (e.g., EMA, FDA and OECD) on AI’s transformative role in accelerating safer, more ethical chemical and pharmaceutical risk assessments.
To ensure diversity and inclusivity, the invited speakers for this session are gender-balanced, with an equal representation of female and male experts.
Speakers
- Anna Vlot – Towards Animal-Free, Human-Centric Toxicology: Cross-Species Transcriptomics and AI-Driven Mechanistic Interpretation
- Marc Teunis – Practical Applications of language models in the toxicological context.
- Swapnil Chavan – A NEURO-SYMBOLIC AI APPROACH FOR PREDICTING BLOOD-BRAIN BARRIER PENETRATION TO SCREEN FOR POTENTIAL NEUROTOXICANTS
- Julia Matyjasiak – AI-POWERED VIRTUAL CONTROL GROUPS: REDUCING ANIMAL USE IN PRECLINICAL SAFETY ASSESSMENT