For this approach I wanted to test the tool Protein Hunter exclusively as there has not been a lot of wet lab validation. I was intentional not to select any hotspots during generation. Designs were in the minibinder range 50 to 100aa for small size but good surface area. Many designs bind the unstructured region of RBX1 while others target the ordered region, and some target both. These were sorted by ipsae values and filtered by K+E+A sequence ratios.