How is AI used to predict drug-induced liver injury in toxicogenomics?

AI is being used to predict drug-induced liver injury (DILI) by finding patterns in toxicogenomic data. Toxicogenomics means studying how drugs change gene activity in liver cells. Standard lab tests often miss DILI because it is unpredictable and can have complex causes ¹⁷. AI models can analyze large amounts of gene data to spot warning signs earlier than older methods ¹⁹. The short answer: AI helps turn gene-level data into predictions of liver risk, and some models are already showing useful accuracy ⁸.

What the research says

Toxicogenomics has become a main driver of AI progress in DILI research. Early computational work used clustering and pathway analysis to find molecular signatures of early liver injury. Later, supervised machine learning allowed researchers to select key features, refine signatures, and validate predictions more rigorously. More recently, deep learning, biologically informed network designs, and generative AI have pushed the field further, helping integrate different types of evidence and improve mechanistic understanding ¹. A 2021 review notes that toxicogenomic studies have moved toward stem cell derived organoids and primary human hepatocytes in 3D models, which help separate toxic from nontoxic compounds and support personalized risk analysis using cells from different donors ⁹.

What to ask your doctor

• How does my liver function get monitored while I take this medicine?
• Are there any gene or blood tests that could show my personal risk of liver injury from this drug?
• What symptoms of liver problems should I watch for, and when should I call you?
• If I have a liver condition like fatty liver, does that change my risk with this medicine?
• Are there any drug interactions that could raise my liver risk?