AI maps hidden pathways of toxic chemical harm in new safety model

Researchers have developed a machine-learning system that automatically generates networks of how chemicals damage living organisms, using hormone-disrupting substances as a test case. The approach could accelerate how regulators and pharmaceutical companies assess chemical safety, potentially reducing testing time and costs while improving prediction accuracy.

Originaltitel: Development of a data-driven approach to Adverse Outcome Pathway network generation: a case study on the EATS-modalities