Deep Learning for Drug-Target Interaction Prediction: A Comprehensive Review

Drug-target interaction (DTI) prediction plays a vital role in drug discovery. However, traditional experimental methods are often time-consuming and resource-intensive. Recently, deep learning (DL) approaches have emerged as powerful and efficient tools for predicting DTIs. This paper provides a structured overview of these DL-based methods, beginning with a review of feature representation strategies for drugs and proteins, followed by a summary of commonly used datasets and evaluation metrics. The review critically examines various DL architectures, including deep neural networks (DNNs), recurrent neural networks (RNNs), convolutional neural networks (CNNs), graph neural networks (GNNs), and Transformer-based models. Furthermore, we discuss their applications in drug repositioning, drug design, and precision medicine. Finally, we address key challenges such as data scarcity and model interpretability, and highlight future research directions including self-supervised learning and explainable artificial intelligence. This review aims to provide a rigorous synthesis of current advances to inform future developments in DL-based DTI prediction.