Home›Drug Pipeline› AI and Machine Learning Reshape Drug-Induced Liver Injury Toxicogenomics
AI and Machine Learning Reshape Drug-Induced Liver Injury ToxicogenomicsArtificial intelligence helps predict liver damage from certain medications
Frontiers in MedicinePublished September 13, 2026DOI ↗Editorial oversight: Dr. Julia Lee, PhD · Oncology, Genomics & Drug Development
AI-generated summary of the cited source, checked by automated accuracy review.
How we work
Share
Key Takeaway
AI and machine learning are advancing DILI toxicogenomics from descriptive profiling to mechanism-driven predictive models.
A systematic review examined how artificial intelligence and machine learning have evolved within drug-induced liver injury (DILI) toxicogenomics. The analysis traces a clear progression from early exploratory work to sophisticated predictive modeling.
Early computational toxicogenomics relied on exploratory approaches such as clustering, time-series analysis, co-expression networks, and basic pathway analyses. These methods aimed to identify molecular signatures of nascent liver injury but remained largely descriptive.
The introduction of supervised machine learning enabled more robust predictive modeling. This phase brought systematic feature selection, signature refinement, and rigorous validation, strengthening the reliability of identified biomarkers.
More recent advances include deep learning, biologically informed networks, and generative AI. These approaches enhance mechanistic interpretation by identifying biologically relevant signatures, integrating multimodal evidence, and reinforcing evidence for established DILI mechanisms such as oxidative stress, mitochondrial dysfunction, altered xenobiotic metabolism, inflammation, and cell death.
The review describes a field transitioning from descriptive profiling toward mechanism-driven predictive toxicology. No specific funding sources or conflicts of interest were reported, and the review did not report safety outcomes, follow-up data, or formal certainty assessments.
How this fits prior evidence
This review extends prior coverage of computational liver injury prediction. Earlier work on risk prediction models for anti-tuberculosis drug-induced liver injury reported a pooled AUC of 0.81 but noted high bias and heterogeneity limiting prophylaxis guidance. In silico modeling of acetaminophen overdose highlighted metabolic parameter importance without clinical outcome data, and in silico fezolinetant hepatotoxicity predictions in MAFLD required cautious interpretation pending clinical data. This review similarly describes methodological progress in DILI toxicogenomics without reporting pooled performance metrics or clinical validation, consistent with the field's ongoing gap between computational prediction and clinical utility.
When a person takes a medication that causes liver injury, it can be hard for doctors to predict exactly how the body will react. This type of injury, known as DILI, is a major concern in medicine. Researchers are now using artificial intelligence and machine learning to better understand these reactions.
Early computer models focused on finding basic patterns in how cells react to new drugs. These early steps helped identify the first signs of liver stress. More recent systems use supervised machine learning to create more accurate models. These tools help experts pick out specific features and refine the data to predict liver damage more reliably.
Newer technologies like deep learning and generative AI are taking things further. Instead of just spotting a problem, these tools help scientists understand the underlying biology, such as how a drug might cause inflammation or stress the cell's energy centers. While these tools are still evolving, they are moving the field from simple observation toward a deeper understanding of how drugs affect the liver.
What this means for you:
Artificial intelligence helps researchers identify the specific biological mechanisms that cause liver damage from drugs.
Common questions
How is artificial intelligence used to study liver damage?
Artificial intelligence and machine learning help researchers move from simple descriptions of liver injury to understanding the actual biological mechanisms. These tools can identify molecular signatures, select key features, and refine predictive models to see how drugs affect the liver.
What specific liver issues can these computer models identify?
Advanced models can help identify several key biological issues, including oxidative stress, mitochondrial dysfunction, and problems with how the body processes chemicals. They also help track inflammation and cell death to better understand how a drug might cause liver injury.
Drug-induced liver injury (DILI) is a critical safety issue in drug development, characterized by its idiosyncratic nature, complex mechanisms, and poor predictability in standard preclinical models. High-dimensional omics strategies, particularly toxicogenomics, have attracted increased interest in addressing the complexity of hepatotoxicity, especially in the context of emerging artificial intelligence (AI) technologies. This review traces the evolution of AI and machine learning (ML) within DILI-related omics research, highlighting toxicogenomics as a primary driver of advancement in this field. We first explore the early studies in computational toxicogenomics, which primarily focused on exploratory approaches, utilizing clustering, time-series, co-expression, and basic pathway analyses to identify molecular signatures indicative of nascent liver injury. We then examine how the adoption of supervised machine learning enabled robust predictive modeling, facilitating systematic feature selection, signature refinement, and rigorous validation. More recently, the field has been further transformed by deep learning, biologically informed network architectures, and generative artificial intelligence. Across these methodological eras, AI has enhanced mechanistic interpretation by identifying biologically relevant signatures, integrating multimodal evidence, and strengthening evidence for established DILI mechanisms, including oxidative stress, mitochondrial dysfunction, altered xenobiotic metabolism, inflammation, and cell death. Ultimately, the synergy of AI and DILI toxicogenomics has transitioned the discipline from descriptive profiling toward mechanism-driven predictive toxicology.
