DeepCyte launched DeeTox Atlas, a single-cell metabolomic reference map of drug toxicity mechanisms intended to support toxicity interpretation and earlier liability detection. The company said the dataset is built from two independent single-cell metabolomics perturbation studies spanning roughly 100 toxicants, about 300,000 cells, and hundreds of metabolites per cell. DeepCyte positions the atlas as foundation data to train predictive models that reduce the need to generate new wet-lab toxicity readouts for each compound. The company also highlighted a mapping to toxicity-mechanism hierarchies anchored to established adverse outcome pathways. In practical terms for drug developers, the near-term value proposition will depend on how well the platform predicts mechanism-level liabilities for novel molecules and whether adoption translates into faster decision-making during lead optimization and candidate selection.