Hannes Stärk

A geometric deep-learning model is faster and more accurate than state-of-the-art computational models, reducing the chances and costs of drug trial failures.