Abstract
Graph Artificial Intelligence (Graph AI) is reshaping biomedical research by providing a powerful framework to integrate, represent, and analyze complex, heterogeneous, and multi-modal datasets. Unlike traditional learning paradigms, graph-based models explicitly capture the relational and structural dependencies inherent to biological systems, enabling deeper insights into disease mechanisms, therapeutic discovery, and precision medicine. In this talk, I will present recent innovations in multi-modal graph learning, transfer learning across heterogeneous domains, and representation learning under noisy and incomplete data. Case studies will demonstrate applications in robust drug response prediction, graph-based denoising of noisy biomedical data, and inference from incomplete datasets. Collectively, these advances illustrate how Graph AI can address challenges in representation learning, domain adaptation, and robust inference, while driving forward the next generation of biomedical discovery.
For more info, please follow this link.