APPLICATION OF AI IN PREDICTING DRUG-TARGET INTERACTION

Drug–target interactions (DTIs) are central to modern drug discovery, guiding the identification of therapeutic mechanisms, safety profiles, and opportunities for drug repurposing. Traditional experimental and computational methods, while valuable, are limited by scalability, cost, and their ability to capture complex biological processes. Recent advances in artificial intelligence (AI), particularly machine learning (ML), deep learning (DL), and graph neural networks (GNNs), have revolutionized DTI prediction by enabling accurate, scalable, and data-driven approaches. AI applications extend across virtual screening, drug repurposing, side-effect prediction, and pharmacokinetic modeling, thereby accelerating the entire drug discovery pipeline. Benchmark resources such as DrugBank, BindingDB, and ChEMBL support model training and evaluation, while explainable AI (XAI) enhances interpretability and trust in predictive outcomes. Despite these advances, challenges persist, including limited data quality, dataset imbalance, poor interpretability of complex models, and computational demands. This review consolidates current knowledge, highlights state-of-the-art AI methodologies for DTI prediction, and identifies future directions such as integration of 3D structural data, generative AI for molecule design, and the adoption of multi-modal and explainable models. Together, these advancements underscore AI’s growing role in achieving efficient, cost-effective, and precision-driven drug discovery.