Combining machine-learning and physics-based workflows to map and identify promising binding sites for binder design. Tools that we used to design our binders, comprise BindCraft, Protein MPNN, Caliby, Boltz2, RF-diffiusion and molecular dynamics simulations.