Custom-designed proteins that can control cells and activate them in a targeted way: Prof. Thomas Schlichthärle wants to use them to advance the development of new medicines. In this new episode of “NewIn,” he explains how researchers in other disciplines can benefit from his methods.

Imagine being able to dock tiny machines onto a stem cell – an as-yet undifferentiated cell of the human body. Those machines would reprogram the cell to transform – into a heart cell, for example, or a kidney cell. What sounds like science fiction is already reality in Thomas Schlichthärle’s lab. The Professor of AI-Guided Protein Design uses artificial intelligence to develop such nanomachines – proteins that do not exist in nature. 

Professor Schlichthärle’s specialty is de novo protein design. In this approach, amino-acid sequences are designed from scratch so that they reliably fold into a very specific 3D structure and perform a specific function. To select the most promising candidates from the theoretically infinite number of possible sequences, Schlichthärle uses machine-learning methods known as deep-learning algorithms. “With artificial intelligence, we can evaluate a much larger number of possible protein sequences much faster. That opens up possibilities we never imagined.” 

Solutions for tomorrow’s medicines 

The proteins developed by him and his team are used, for example, in organoid research. They can reprogram stem cells so they adopt a very specific state and, for instance, develop into heart, kidney, or pancreatic cells. This makes it possible to produce organoids – “mini-organs” made from human cells – in a targeted way. These help researchers study diseases more realistically and test novel drugs. 

The protein “neotrophin,” which he has newly developed and  which binds to the TrkA receptor on the surface of specialized nerve cells, also has potential for future medicines. In injuries to the peripheral nervous system, it could act as a growth factor, helping nerve cells regenerate without triggering the pain that is usually associated with this. “If we understand the molecular mechanisms by which proteins act in and on cells, we can also control which signaling pathways in the cell should be activated,” Schlichthärle says. In the case of “neotrophin,” it is possible to only activate cell growth and leave out the signaling pathway for pain. A start-up in Seattle wants to further develop this compound together with the researcher and bring it to market as a medicine. 

A fascination with future technologies has been with Thomas Schlichthärle for a long time. He loves Isaac Asimov’s science-fiction novels, is interested in bio- and nanotechnology, and enjoys programming. He studied molecular medicine in Tübingen and molecular bioengineering at TU Dresden. “I wanted to understand exactly how the human body works and how new methods can be developed to tackle disease.” 

Schlichthärle wrote his master’s thesis at the Wyss Institute at Harvard University in Boston, which is dedicated to translating research results into industry. For his doctorate, he returned to Germany – to the Max Planck Institute of Biochemistry in Martinsried near Munich. There, he worked on super-resolution microscopy and computer-assisted methods for data analysis. A central problem at the time: To be visible under the microscope, the proteins being studied have to be tagged with so-called labels. Antibodies are usually the probes of choice. However, optimizing them for individual binding pockets – docking sites for other molecules on the surface of cells – is labor-intensive. Because of their size, such antibodies also typically have low binding efficiency.

Speeding up protein development

The solution: Instead of antibodies, smaller proteins are used – designed on a computer and optimized specifically for individual binding pockets. “That’s a much faster and more elegant solution,” Schlichthärle says. To learn this method, he went to the Institute for Protein Design at the University of Washington in Seattle as a postdoc, where he worked in the lab of Nobel Prize laureate David Baker. In June 2025, he was appointed to the Professorship of AI-guided Protein Design at TUM. The position was newly created as part of the Hightech Agenda Bavaria and is supported by the Wübben Stiftung Wissenschaft. 

Solving puzzles and tackling problems – that’s what Thomas Schlichthärle enjoys most about his work. With those skills, he wants to advance science, including beyond his own field. “Our methods and research results from protein design are helpful for researchers, but also for start-up teams across many disciplines,” Professor Schlichthärle says. As examples, he mentions developing contrast agents for magnetic resonance imaging or new enzymes for producing sugar substitutes. 

In addition to his research at the TUM Center for Functional Protein Assemblies and the newly founded Center for Smart Drug Design, he wants to wants to establish a Protein Design Accelerator at TUM. This service unit is intended to speed up the development of new proteins and to be available to researchers at TUM as well as external partners. He is also involved in the new Cluster of Excellence Biosystems Design Munich (BiosysteM). There, researchers from the two Munich universities and additional scientific institutes are working on biomolecular machines, intelligent materials, and novel cell systems – bringing the future a little closer. 

About the person

Thomas Schlichthärle studied Molecular Medicine for his bachelor’s degree in Tübingen and Molecular Bioengineering for his master’s degree at TU Dresden. After a research stay at the Wyss Institute at Harvard University in Boston, USA, he completed his doctorate at the Max Planck Institute of Biochemistry in Munich. He then worked as a postdoctoral researcher at the University of Washington in Seattle, where he developed novel synthetic growth factors. In 2025, he was appointed Professor of AI-guided Protein Design at TUM.

Further information and links

• The professorship is funded by Hightech Agenda Bayern (HTA)

Contact about this article:

Prof. Dr. Thomas Schlichthärle
AI-guided Protein Design
Technichal University of Munich
Tel: +49 (89) 289 13349 
E-Mail: thomas.schlichthaerle@tum.de

Technical University of Munich
Corporate Communications Center

Undine Ziller
undine.ziller@tum.de
presse@tum.de
Teamwebsite

Original article: https://www.tum.de/en/news-and-events/all-news/press-releases/details/designing-proteins-for-targeted-cell-activation