Fig. 7.

Computational analysis of the CC structures. a, b Orthogonal views of snapshots from the CpHMD simulations of CC-Type2-LL-L17E (a) and CC-Type2-IL-Sg-L17E (b) at pH 4. The backbone structures are represented by ribbons, Glu-17 with purple sticks, and water molecules and sodium ions (blue) as van der Waals’ spheres. c, d Stepwise pH titration of the Glu-17 cluster in CC-Type2-LL-L17E (c) and CC-Type2-IL-Sg-L17E (d) obtained through CpHMD. From left to right, titration events are ${\left(\mathrm{G}lu17\right)}_x^{1-}$ to ${\left(\mathrm{G}lu17\right)}_x^{2-}$ (yellow), ${\left(\mathrm{G}lu17\right)}_x^{2-}$ to ${\left(\mathrm{G}lu17\right)}_x^{3-}$ (green), ${\left(\mathrm{G}lu17\right)}_x^{3-}$ to ${\left(\mathrm{G}lu17\right)}_x^{4-}$ (turquoise), ${\left(\mathrm{G}lu17\right)}_x^{4-}$ to ${\left(\mathrm{G}lu17\right)}_x^{5-}$ (blue), ${\left(\mathrm{G}lu17\right)}_x^{5-}$ to ${\left(\mathrm{G}lu17\right)}_x^{6-}$ (purple) and ${\left(\mathrm{G}lu17\right)}_x^{6-}$ to ${\left(\mathrm{G}lu17\right)}_x^{7-}$ (pink). Note that the ${\left(\mathrm{G}lu17\right)}_x^0$ to ${\left(\mathrm{G}lu17\right)}_x^{1-}$ (red) transitions are not observed in the pH range modelled. Calculated pK_as are tabulated in Supplementary Tables 8 and 9. e Normalised per-chain BUDE scores from modelling all sequences (rows) onto all structures (columns) in ISAMBARD, with total energy in BUDE points (left) and the extracted steric component (right). For each row, raw scores were normalised relative to the score for the named sequence modelled onto its cognate backbone structure (i.e., the top-left-to-bottom-right diagonal), which is assigned zero and represented by white boxes. In this scheme, red boxes correspond to worse scores and less-favourable pairings of that sequence with the different structures modelled through the row, i.e., if atom–atom clashes appear in the model; and blue boxes represent better scores indicating that the sequence is accommodated better into the alternate structure