view article
Figure 2
Number of predicted sequence-register errors plotted by the predicted residue shift (left), by the number of affected residues (middle) and by the fraction that the affected residues represent compared with the overall structure (right). Errors consisting of a shift of more than ten residues, affecting more than 50 residues or more than 10% of the residues in the structure have been omitted for clarity (127, 248 and 27 errors, respectively). Note that where two errors were separated by three residues or fewer they were combined (see Section 2[link]), meaning that where the two errors involved shifts of different numbers of residues, the average shift for a predicted register error may be non-integral. Note that the numbers here are after sequence-redundancy removal and represent unique errors within and between PDB entries.

IUCrJ
Volume 11| Part 6| November 2024| Pages 938-950
ISSN: 2052-2525
https://doi.org/10.1107/S2052252524009114