Artificial intelligence in pharmaceutical sciences: A comprehensive review

Abstract

The integration of artificial intelligence (AI) and machine learning (ML) into pharmaceutical sciences has catalyzed transformative advancements across drug discovery, clinical development, manufacturing, and post-market surveillance. This review comprehensively examines AI's role in modern pharmacotherapy, beginning with its historical evolution in life sciences and progressing to cutting-edge applications such as AlphaFold-driven protein modeling, natural language processing (NLP) for biomedical literature mining, and AI-augmented pharmacovigilance. Methodologically, we synthesize interdisciplinary insights from peer-reviewed literature (2013–2023), highlighting innovations in cheminformatics (e.g., QSAR, RDKit), predictive toxicology, and personalized medicine. Case studies illustrate AI's capacity to compress drug development timelines, as seen in COVID-19 repurposing efforts and de novo kinase inhibitor design. However, challenges persist, including algorithmic bias, regulatory ambiguities, and the “black-box” nature of deep learning models. By critically evaluating successes and limitations, this review underscores AI's potential to redefine pharmaceutical innovation while advocating for robust frameworks to ensure ethical, transparent, and clinically translatable AI deployment.

Keywords

• Artificial intelligence
• Molecular modeling
• Pharmacovigilance
• Deep learning
• Cheminformatics
• Drug discovery