On May 5, 2026, Prof. Jan Baumbach (Director of Computational Systems Biology, University of Hamburg) gave a lecture as part of our seminar on Recent Methods in Medical Informatics. The talk was titled “Network Medicine GPT – A Foundation Model for Disease Mechanism Mining and Drug Repurposing.”

The presentation focused on NetMedGPT, a new AI model designed to analyze large-scale medical data, identify relationships between diseases and drugs, and thereby reveal potential new therapeutic approaches. Unlike many previous models, NetMedGPT can handle multiple tasks simultaneously, such as predicting which compounds may be effective for specific diseases, identifying possible side effects or contraindications, and even evaluating the use of drugs for officially unapproved indications (“off-label use”).

A particularly interesting aspect of the model is its generative capability, which enables the creation of biologically plausible networks that can provide new insights into disease mechanisms. In this way, NetMedGPT supports researchers in developing new hypotheses and potentially accelerating drug repurposing.

For those interested, an interactive interface is available that allows users to query the model’s predictions directly using natural language: https://apps.cosy.bio/netmedgpt/.

The lecture clearly demonstrated how modern AI methods can advance medical research and, in the long term, support personalized medicine.