Role of Artificial Intelligence in bioinformatics: Revolutionizing molecular docking and DNA tokenization

Bioinformatics has become a crucial discipline that connects biology and computational analysis, extracting meaningful insights from complex biological datasets. Traditional bioinformatics approaches, which rely on statistical models and manual analysis, are often inadequate for handling such complexity. The integration of artificial intelligence (AI) and machine learning (ML) has revolutionized bioinformatics by providing powerful computational tools for solving complex problems in biosciences. AI-based models automate complex tasks by improving data accuracy, minimizing noise, and revealing hidden biological patterns. In particular, ML has transformed molecular screening and structure-based drug discovery by reducing dependency on predefined biological targets, thereby streamlining the drug discovery pipeline. Recent advancements in deep learning (DL), transformer architecture, diffusion models, and foundation models have significantly improved molecular docking, protein structure prediction, virtual screening, and genomic sequence analysis. AI-driven docking frameworks such as DiffDock, GNINA, FeatureDock, and AlphaFold have enhanced the prediction of binding poses, affinities, and protein structures with improved accuracy and computational efficiency. Similarly, transformer-based genomic models such as DNABERT, DNABERT-2, BigBird, and Enformer have revolutionized DNA tokenization and long-range sequence analysis. This review highlights the major contributions of AI and ML in bioinformatics, with particular focus on their applications in molecular docking and DNA tokenization-based genomic analysis. Furthermore, the review discusses current challenges and limitations, while emphasizing the future potential of hybrid and biologically informed AI frameworks in next-generation bioinformatics.