Artificial Intelligence in Healthcare and Drug Discovery: From Predictive Diagnostics to Precision Medicine

Abstract

The integration of artificial intelligence (AI) and deep learning with the biomedical research marks a critical inflection point, shifting the paradigm from addressing symptoms to predicting and acting at the molecular level. In many diagnostic models, the chain reaction triggered by a chronic infection is often overlooked. In this review, we examine the capability of AI models to unravel these complex dynamics within viral-host systems. By analyzing 39 key studies, we evaluate the application of novel computational methods in two major domains: advanced clinical diagnostics and de novo drug design. Silently existing chronic conditions, such as chronic Hepatitis, can lead to immune exhaustion, leaving individuals more susceptible to secondary infections including SARS-CoV-2 (COVID-19) and Human Papillomavirus (HPV). For diagnostics, we focus on genomic surveillance, radiomics, and Explainable AI (XAI) for identifying multi-organ pathologies. For pharmacology, we examine the transformative role of generative models, including AlphaFold 3, HelixFold 3, and large foundational models (Evo 2, Geneformer, ESM-2). We discuss Graph Neural Networks (GraphDTA) and diffusion models in successfully identifying novel antimicrobial agents like Halicin and Abaucin. Finally, we explore clinical and regulatory barriers to the full integration of AI in mapping patient pathologies.