Overview

Axiom Bio uses machine learning and biochemistry to predict drug-induced liver toxicity. Its platform analyzes a molecule’s structure with AI models trained on the world’s largest proprietary dataset of human toxicity data, derived from screening over 115,000 small molecules on primary human liver cells and paired with pharmacokinetic and clinical outcome data. The system can reveal mechanisms behind toxicity, such as mitochondrial disruption, ER stress, and reactive oxygen species formation, and it outputs risk assessments for hepatotoxicity to help researchers de-risk drug development. The service is designed to be cheaper and faster than lab testing—costing roughly $100–$450 per compound compared with $3,000–$15,000 for traditional experiments—and it aims to replace animal testing with in-silico predictions. Axiom’s goal is to enable safer, more efficient drug development by providing accurate, mechanism-informed toxicity predictions to pharmaceutical and biotech clients.