At the 10th International Summer School on AI and Big Data, Prof. Florian Praetorius will talk about Protein design in the age of AI.
Computational protein design can be use do generate new proteins from scratch (de novo design) or to modify structure and properties of natural proteins. Over the past few years, machine learning based methods have revolutionized the field, be it for backbone generation, sequence design and optimization, or computational filtering of designs. Here, I will give a non-exhaustive overview of recent developments in computational protein design and present some examples of design projects that illustrate the strengths and weaknesses of the current methods.
Florian Praetorius did his PhD in Munich at TUM with Hendrik Dietz, working on DNA origami. He then moved to Seattle to learn computational protein design in the lab of David Baker. In his own lab at ISTA, Florian aims to combine de novo protein design with DNA nanotechnology.