This is a two-sided design downside, the place we will design the sequence of each binding companions. The purpose is to design a pair of proteins that can self-associate in resolution. This is particularly helpful for bioengineering functions the place we’d need to use a brief peptide as a “sign” that may be acknowledged by its binding associate. Although the quick peptide could also be versatile in resolution, we want to bind it within the form of beta strand, so the goal spine has been frozen as a beta strand. However, you should still change the sidechains of the strand goal to any amino acid you need!

AlphaFold predictions are enabled for the designed binder. Once you have designed a binder for the strand goal, add your resolution for AlphaFold utilizing the AlphaFold prediction software. AlphaFold will predict the construction of your binder chain solely (i.e. within the unbound state, within the absence of the strand goal). If you load this prediction, then Foldit will try to align the prediction together with your resolution (i.e. within the certain state, making an interface with the strand goal). If you proceed working off of the AlphaFold prediction, you might have to make changes on the interface the place the binder interacts with the goal. See the puzzle feedback for added Objectives.