International Journal of Scientific Research and Engineering Development

Artificial intelligence (AI) is reshaping drug discovery by learning patterns from chemical, biological, and clinical data to propose targets, design molecules, and predict safety and efficacy. This review synthesizes recent advances across the drug discovery and development pipeline, emphasizing modern deep learning (DL) methods such as transformers and diffusion models for de novo design, graph neural networks for molecular property prediction, and protein structure prediction systems (e.g., AlphaFold2) that expand structure-enabled screening. We provide a pragmatic taxonomy of AI tasks, data modalities, and validation practices; highlight representative industrial and regulatory milestones; and discuss limitations including data bias, assay shift, uncertainty quantification, interpretability, and governance. Finally, we outline a riskbased framework for deploying AI in regulated settings and identify research directions—multimodal foundation models, active learning with laboratory feedback, and human-centered design—needed to convert algorithmic novelty into reproducible clinical impact.

Nishchith Gosala J, Pooja K N,"Artificial Intelligence-Based Approaches in Drug Discovery and Development: A Review" International Journal of Scientific Research and Engineering Development, V9(2): Page(280-282) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.