From Artificial Intelligence to Network Medicine: New ways to understand and treat diseases

Summary

Artificial Intelligence (AI) is transforming biomedicine by enabling data-driven approaches to disease understanding, clinical decision support, and therapeutic discovery. This seminar provides an overview of AI methodologies applied to biomedical data, from classical machine learning to deep learning, foundation models, and large language models, highlighting both their potential and their limitations in healthcare contexts. The talk focuses on drug repurposing as a key application, illustrating how AI and network medicine can jointly model diseases, drugs, and molecular mechanisms as interconnected systems. Network-based representations and proximity measures are introduced to identify repurposing opportunities, followed by recent advances using Graph Neural Networks to predict novel disease–drug associations from heterogeneous biomedical knowledge. Through selected case studies, the seminar demonstrates how integrating AI with network medicine offers a scalable and mechanistic framework to accelerate drug discovery and improve our understanding of complex diseases.