Figure 1. Computational workflow for Complete Combinatorial Mutational Enumeration Interface computation:

Interface residues are computed on the ACE2/L strain RBD complex. Computational Protein Design: the global minimum energy conformation is computed. Sequence enumeration: all sequences within a 8 kcal/mol threshold are enumerated. Side chain positioning: the energy of all sequences enumerated on the complex form is computed on the RBD apo form. Next, filters are applied in order to only keep sequences with stable apo conformation and good affinity towards ACE2. Fitness Landscape Analysis: the $\mathrm{\Delta }\mathrm{\Delta }\mathrm{G}$ mutational fitness landscape is analyzed and 3,272 local optima sequences are retained.