Skip to main content
Loading

From signal detection to causal prediction: advancing AI for rare and severe adverse drug reactions

09 Sept 2026
Drug Response Prediction
  • Causality from Real-World Data
    How can we reliably distinguish true drug-induced effects from confounding and bias in FAERS and EHR data?

  • Predicting Rare Toxicities
    What approaches actually work for predicting rare but severe events like DILI and cardiotoxicity—and where do current models fail?

  • From Signals to Mechanisms
    Can integrating multimodal data (clinical, omics, chemical) help move from statistical signals to mechanistic understanding of toxicity?

  • Validation and Trust
    What does robust validation look like for ADR prediction models across populations and real-world settings?

  • Actionability of AI Predictions
    What level of predictive or causal evidence is sufficient to influence clinical, regulatory, or portfolio decisions?

Industry Expert
Bamboo Lin, Safety Data Science Director - Genentech