Predictive Classification of Clinical Drug Toxicity Using Molecular and Pharmacological Characteristics
DOI:
https://doi.org/10.69980/wmh06927Keywords:
Drug Toxicity, Drug-Induced Liver Injury, Predictive Classification, Pharmacological Characteristics, Molecular DescriptorsAbstract
Drug-induced liver injury is still a significant problem in drug development as clinically important toxicity can occur following exposure even when preclinical results are satisfactory. In this study, the predictive classification of clinical drug toxicity was examined using the drug-induced liver injury severity and toxicity (DILIst) dataset from the U.S. Food and Drug Administration (FDA). A dataset of 1,279 distinct drugs that were DILIst-positive or DILIst-negative, along with route-of-administration data, was provided. Descriptive analyses described prevalence of toxicity and patterns of pharmacological routes and chi-square testing assessed the association between categories of routes and toxicity classification. The route-based logistic regression model was tested by stratified 5-fold cross validation and an independent 20% test set. Of 1,279 drugs, 768 (60.0%) were DILIst-positive and 511 (40.0%) were DILIst-negative. Route of administration was reported for 1013 compounds and most were administered orally. A significant difference in the proportion of DILIst-positive compounds was detected between the route categories, but the ability to discriminate with route alone was limited. The analysis confirms the utility of curated clinical toxicity labels, and reveals that pharmacological route data alone is not sufficient to predict toxicity well. For future studies and validation, integration of validated molecular descriptors, structural fingerprints, exposure variables and mechanistic pharmacology is therefore needed to improve and interpret clinical drug-toxicity classification models.
References
1. Aleo, M. D., et al. (2020). Moving beyond binary predictions of human drug-induced liver injury (DILI) toward contrasting relative risk potential. Chemical Research in Toxicology, 33(1), 223-238. https://doi.org/10.1021/acs.chemrestox.9b00262
2. Allison, R., et al. (2023). Drug induced liver injury-A 2023 update. Journal of Toxicology and Environmental Health, Part B, 26(8), 442-467. https://doi.org/10.1080/10937404.2023.2261848
3. Ancuceanu, R., et al. (2020). Computational models using multiple machine learning algorithms for predicting drug hepatotoxicity with the DILIrank dataset. International Journal of Molecular Sciences, 21(6), 2114. https://doi.org/10.3390/ijms21062114
4. Cao, A., Zhang, L., Bu, Y., & Sun, D. (2024). Machine learning prediction of on/off target-driven clinical adverse events. Pharmaceutical Research, 41(8), 1649-1658. https://doi.org/10.1007/s11095-024-03742-x
5. Cavasotto, C. N., & Scardino, V. (2022). Machine learning toxicity prediction: Latest advances by toxicity end point. ACS Omega, 7(51), 47536-47546. https://doi.org/10.1021/acsomega.2c05693
6. Chierici, M., et al. (2020). Predictability of drug-induced liver injury by machine learning. Biology Direct, 15(1), 3. https://doi.org/10.1186/s13062-020-0259-4
7. Guan, J., et al. (2025). StackDILI: Enhancing drug-induced liver injury prediction through stacking strategy with effective molecular representations. Journal of Chemical Information and Modeling, 65(2), 1027-1039. https://doi.org/10.1021/acs.jcim.4c02079
8. Guo, W., et al. (2023). Review of machine learning and deep learning models for toxicity prediction. Experimental Biology and Medicine, 248(21), 1952-1973. https://doi.org/10.1177/15353702231209421
9. Hao, Y., Romano, J. D., & Moore, J. H. (2022). Knowledge-guided deep learning models of drug toxicity improve interpretation. Patterns, 3(9), 100565. https://doi.org/10.1016/j.patter.2022.100565
10. Jaganathan, K., Tayara, H., & Chong, K. T. (2021). Prediction of drug-induced liver toxicity using SVM and optimal descriptor sets. International Journal of Molecular Sciences, 22(15), 8073. https://doi.org/10.3390/ijms22158073
11. Khan, M. Z. I., et al. (2024). Comprehensive hepatotoxicity prediction: Ensemble model integrating machine learning and deep learning. Frontiers in Pharmacology, 15, 1441587. https://doi.org/10.3389/fphar.2024.1441587
12. Li, T., Tong, W., Roberts, R., Liu, Z., & Thakkar, S. (2021). DeepDILI: Deep learning-powered drug-induced liver injury prediction using model-level representation. Chemical Research in Toxicology, 34(2), 550-565. https://doi.org/10.1021/acs.chemrestox.0c00374
13. Liu, A., Seal, S., Yang, H., & Bender, A. (2023). Using chemical and biological data to predict drug toxicity. SLAS Discovery, 28(3), 53-64. https://doi.org/10.1016/j.slasd.2022.12.003
14. Liu, X., et al. (2020). Machine-learning prediction of oral drug-induced liver injury (DILI) via multiple features and endpoints. BioMed Research International, 2020, 4795140. https://doi.org/10.1155/2020/4795140
15. Martin, M. T., et al. (2022). Early drug-induced liver injury risk screening: “Free,” as good as it gets. Toxicological Sciences, 188(2), 208-218. https://doi.org/10.1093/toxsci/kfac054
16. Mostafa, F., Howle, V., & Chen, M. (2024). Machine learning to predict drug-induced liver injury and its validation on failed drug candidates in development. Toxics, 12(6), 385. https://doi.org/10.3390/toxics12060385
17. Norman, B. H. (2020). Drug induced liver injury (DILI): Mechanisms and medicinal chemistry avoidance/mitigation strategies. Journal of Medicinal Chemistry, 63(20), 11397-11419. https://doi.org/10.1021/acs.jmedchem.0c00524
18. Rao, M., et al. (2023). AI/ML models to predict the severity of drug-induced liver injury for small molecules. Chemical Research in Toxicology, 36(7), 1129-1139. https://doi.org/10.1021/acs.chemrestox.3c00098
19. Sharma, B., et al. (2023). Accurate clinical toxicity prediction using multi-task deep neural nets and contrastive molecular explanations. Scientific Reports, 13, 4908. https://doi.org/10.1038/s41598-023-31169-8
20. Teng, S., et al. (2023). MolFPG: Multi-level fingerprint-based graph Transformer for accurate and robust drug toxicity prediction. Computers in Biology and Medicine, 164, 106904. https://doi.org/10.1016/j.compbiomed.2023.106904
21. Tran, T. T. V., Wibowo, A. S., Tayara, H., & Chong, K. T. (2023). Artificial intelligence in drug toxicity prediction: Recent advances, challenges, and future perspectives. Journal of Chemical Information and Modeling, 63(9), 2628-2643. https://doi.org/10.1021/acs.jcim.3c00200
22. U.S. Food and Drug Administration. (2023). Drug-Induced Liver Injury Severity and Toxicity (DILIst) Dataset. FDA Liver Toxicity Knowledge Base. https://www.fda.gov/science-research/liver-toxicity-knowledge-base-ltkb/drug-induced-liver-injury-severity-and-toxicity-dilist-dataset
23. Vall, A., et al. (2021). The promise of AI for DILI prediction. Frontiers in Artificial Intelligence, 4, 638410. https://doi.org/10.3389/frai.2021.638410
24. Wang, L., et al. (2020). An overview of machine learning and big data for drug toxicity evaluation. Chemical Research in Toxicology, 33(1), 20-37. https://doi.org/10.1021/acs.chemrestox.9b00227
25. Wang, M. W. H., Goodman, J. M., & Allen, T. E. H. (2021). Machine learning in predictive toxicology: Recent applications and future directions for classification models. Chemical Research in Toxicology, 34(2), 217-239. https://doi.org/10.1021/acs.chemrestox.0c00316
26. Zhang, H., Zhang, H.-R., Hu, M.-L., & Qi, H.-Z. (2022). Development of binary classification models for assessment of drug-induced liver injury in humans using a large set of FDA-approved drugs. Journal of Pharmacological and Toxicological Methods, 116, 107185. https://doi.org/10.1016/j.vascn.2022.107185
