The liver takes the brunt of the job clearing drugs from the body. Now, researchers have developed a computational model that predicts which drug candidates will prove toxic to the organ (Chem. Res. Toxicol. 2016, DOI: 10.1021/acs.chemrestox.5b00465). Denis Mulliner of Sanofi and his team developed a model that incorporated human and animal toxicity data from 3,712 compounds—three or more times the number used in most earlier models—organizing the data based on each chemical’s molecular properties and mechanism of liver toxicity. The model looks for common chemical and structural properties that lead to a particular type of toxicity. The team then tested its model with 269 proprietary compounds not included in the database. The liver toxicity of these compounds had been tested in animals, and the new model correctly identified 72% of the hepatotoxic compounds. The researchers are sharing this database, along with the model’s source code, with the scientific community “to advance the field of predictive toxicology,” Mulliner says.