Figure 8
Venn diagram showing the number of residues found in modelling errors that could only be detected using the CMO approach (turquoise), the support-vector machines classifier (SVM, red), both methods (intersection) or neither of them (outside the circumferences). Residues found in the training set for the SVM classifier were discarded for this analysis. The numbers of residues involved in non-register-related modelling errors are shown in parentheses. |