Simon Dürr studied Life Science at the University of Konstanz with stays at Université de Montreal, Kings College London, Uppsala University and EPFL.
He completed his PhD at EPFL and Stanford University working on design of novel metalloproteins using deep learning, evolutionary algorithms, and molecular modelling. He has published state of the art deep learning models for modelling metal cofactors in proteins and used these algorithms to design novel metalloproteins. He has also published various articles about computational modelling of proteins using classical methods (e.g. molecular dynamics simulations) and high-level quantum mechanical molecular mechanical (QM/MM) simulations to study ligand binding, tunnel formation and enzymatic catalysis.
He collaborates with the startups HuggingFace and AdaptyvBio on machine learning for proteins. He built popular FAIR web applications for some of the workhorse models of modern protein design such as ProteinMPNN. To further the cause of FAIR scientific software he developed several open source packages such as a set of open components for visualization of proteins and small molecules for the Gradio framework.
Simon Dürr also runs the largest open platform of open and FAIR scientific illustrations: bioicons.com.
Simon Dürr’s current research interests include:
- Computational optimization of biocatalysts using molecular modelling and deep learning
- De novo design of functional proteins
- Lab-in-the-loop approaches and use of AI agents for lab automation
- Scientific data visualization and scientific illustrations using web technologies
More information on simonduerr.eu
SWITCH edu-ID
Administration
