In a presentation entitled “AI-guided Protein Design for Biomedical Applications: Engineering growth factor mimetics for stem cell differentiation and neuronal regeneration”, Prof. Thomas Schlichthärle introduced his current research to over a dozen fellow professors and a room packed with students at the Center for Protein Assemblies (CPA) at Campus Garching.  

The event began with a brief introduction by Prof. Andreas Bausch, Director of the CPA, who outlined Prof. Schlichthärle’s education and research career. His scientific education started at Eberhard Karls University of Tübingen, where he received a BSc in molecular medicine, followed by an MSc in molecular bioengineering from the Technical University Dresden. For his doctorate, Schlichthärle attended Ludwig Maximilian University of Munich and the Max Planck Institute for Biochemistry. He came to TUM in June 2025 after almost 5 years at the Institute for Protein Design at the University of Washington (USA). 

Schlichthärle’s lab at TUM develops AI-driven methods for protein design to precisely engineer biological systems, from molecular discovery to cellular reprogramming. Treating biology as a programmable system, they use machine learning and generative models to create synthetic proteins, rewire signaling pathways, and build novel cellular functions. Their long-term goal is to integrate artificial intelligence with synthetic biology to predict and design life at molecular, cellular, and network scales for real-world applications.  

Prof. Schlichthärle warmly thanked Prof. Bausch for his introduction and the support of the CPA for research infrastructure, which makes his research possible. In the presentation, he focused on receptor tyrosine kinases, which are activated by growth factors and control cell growth, survival and cellular differentiation. In particular, he showed the de novo design of fibroblast- and nerve growth factor mimetics that can be used for highly-specific stem cell differentiation pathways and neuronal regeneration. The presentation also gave an overview of current deep-learning based tools for accelerated protein engineering that allow researchers to create new-to-nature proteins.  

In conclusion, Schlichthärle remarked, “What I really want you to take from this talk is that protein design in the age of machine learning has gotten much, much easier.”   

Following a diverse and active Q&A session, Prof. Schlichthärle received enthusiastic applause. The afternoon concluded with snacks and engaging discussions in the CPA foyer. 

Further information and links

• Prof. Thomas Schlichthärle’s research group AI-Guided Protein Design
• Department of Bioscience
• Center for Protein Assemblies (CPA) 

Press Contact  
communications(at)nat.tum.de  
Team website