Figure 3
From: Probabilistic grammatical model for helix‐helix contact site classification
Performance comparison with standard methods. Average AUC ROC of classifiers in 4‐fold CV (horizontal axis) and when trained in the PDBTM150NR dataset and tested in WDG150NR dataset (vertical axis). Left: unweighted average, right: weighted average of AUC ROC. Sample method designations: gconcat ‐ global HMM/concatated helices scanned together; l2parts best3 ‐ local HMM/2 parts of a helix pair scanned independently; best3 ‐ best classifiers out of 3trained. In the case of the PCFGs, accessibility grammars are used for c1 and c4, and van der Waals grammars are used for c2 and c3.