Cellarity unveiled an integrated multi‑omics and AI framework to predict drug‑induced liver injury (DILI), publishing methods and validation data in Nature Communications. The clinical‑stage company combined transcriptomics, proteomics and computational models to flag hepatotoxicity earlier than conventional in vitro assays. Their framework aims to reduce late‑stage attrition by identifying safety liabilities before human testing and to guide medicinal chemistry optimization. The paper includes performance metrics across diverse compound sets; Cellarity positions the method as a toxicology tool that augments standard safety pharmacology and could be integrated into preclinical pipelines for lead selection and candidate de‑risking.