DSSS - Use of deep learning in protein design

Proteins are the machines of life. They are diverse macromolecules essential for all cellular processes. Nature has generated an impressive set of proteins through evolution. Many of its structures and even more sequences are known by now. This enormous amount of data can be leveraged for protein design. The design of custom proteins such as reagent antibodies, biosensors or enzymes, is a major goal in protein biochemistry and will be necessary to tackle global challenges that we face today. Approaches to design proteins and enzymes are rapidly evolving with the use of AI techniques enabling higher success rates than before. Here I will discuss different approaches including the uee of language models and show some highlights from our work on the design of complex proteins and enzymes.