Colloquium - Zhiyong Lu, "From Genes to the Clinic: AI Agents for Biomedical Discovery and Decision Support"

Abstract: 
The rapid advancement of large language models is enabling a new generation of domain-specific agents capable of reasoning, retrieving, and acting in complex biomedical contexts. In this talk, we present two representative systems—GeneAgent (Nature Methods 2025), an AI agent for self-verified gene set analysis, and AgentMD (Nature Communications 2025, NeurIPS 2024), a clinical reasoning agent for patient-specific risk calculation. Through real-world use cases, we will discuss the design principles behind these biomedical agents, and how such AI agents can support biomedical discovery and clinical practice, as well as challenges and limitations in medical agentic AI research. 

Bio:
Dr. Zhiyong Lu is a tenured Senior Investigator at the NIH/NLM, leading research in biomedical text/image processing, AI, and machine learning. In addition, Dr. Lu is Adjunct Professor at the University of Illinois Urbana-Champaign (UIUC) and Associate Editor for the journal JAMIA and Bioinformatics. Dr. Lu is a highly cited researcher with over 400 publications, and his AI research has been deployed in real-world systems like PubMed & LitCovid, serving millions each day. Dr. Lu’s research has been frequently featured in major news outlets such as Nature and recognized with numerous awards, including the NIH Director’s Challenge Award, the Clinical Center CEO Award, and the NLM Regents Award. Dr. Lu has been elected to the American College of Medical Informatics (ACMI) and the International Academy of Health Sciences Informatics (IAHSI).