Figure 7. Contingency relationships between distinct peptide positions.

(A) Logos of p−1 preferences in 21 different peptide contexts. Seven 4-residue C-flank sequences were appended to each of three core motifs (…[LP]PxY), and then the p−1 codons were randomized and tested for binding YAP^WW1 and NEDD4^WW3. For clarity, the logos show only the most preferred residues, and they omit data from peptides for which the difference in z-score between the tested WW-Cln2 fusion and unfused Cln2 was less than 1.

(B) Logos showing preferences at p−1 when the p+2 residue is varied (left), or preferences at p+2 when the p−1 residue is varied (right).

(C) Logos showing preferences at p+3 or p+4 when the p+1 residue is varied. Data in panels A-C are derived from the median synonym z-scores from 3 independent experiments.

(D) Predicted conformations of peptides bound to YAP^WW1. Bottom, prediction for the UGR2 peptide, showing 3 side chains for reference (−2P, 0Y, +1S). Above, analogous predictions for peptides with p+1 Trp substitutions (+1W), aligned with the bottom structure and hiding the YAP^WW1 domain for clarity. The Trp groups are predicted to lie next to the p0 Tyr groups (0Y) with their planar faces in a perpendicular geometry.