Supplementary Figure 1: Design and characterization of danoprevir:NS3a complex reader libraries.

a, Process of Rosetta re-design-informed design of a combinatorial D5 interface library. b, Enrichment ratios of the DNCR1 site saturation mutagenesis (SSM) library sorted for (positive sort, top) or against (negative sort, bottom) binding to 50 nM NS3a in the presence of 500 nM danoprevir. The color scale has been flipped for the negative sort such that for both sorts, blue corresponds to predicted weaker binders, and red corresponds to predicted tighter binders. Gray boxes with letters are the wild-type residue and other gray boxes are positions with <15 counts in the naïve library sequencing results. c, Sequence logos of the theoretical library for the second combinatorial library varying the DNCR1 interface (top), and the mutations found in the final enriched clones (bottom). Residue identities at the varied positions are indicated for the starting DNCR1 and final DNCR2. d, Progression of binding improvement from DHR79 to D5 to DNCR1 to DNCR2 as measured by the deviation from average enrichment ratio of the DNCR1 SSM values at each position. Gray shaded region indicates the range of enrichment ratios of all amino acids at each position, and vertical gray bars indicate positions at the interface.