Some protein surfaces are flat and hard to bind. We used a structure model to propose binders for them.

A first round of designs was tested in the lab, and the results guided a second round.

We report what worked, what did not, and what we changed.