AI design of protein binders targeting hydrophilic sites is difficult, so Sappington et al. conditioned RFdiffusion to form β-strands which line up with hydrophilic interfaces. Another useful tool which broadens the applicability of AI protein design!
This study demonstrates the capability of deep learning protein design models in generating functionally validated β-strand pairing interfaces, expanding the structural diversity of de novo binding proteins and accessible target surfaces.
