Figure 1.

ROC curves of structure pair retrieval on the fold level dataset. To determine corresponding residues in protein pairs, TM-align (the left column), HHpred (the middle column), and SUPRB (the right column) were used. Three different bases for residue contact definitions are used, the Cα-Cα distance (the first row), the Cβ-Cβ distance (the second row), and the heavy atom distance (the third row). A, ROC curves using TM-align alignments and the Cα-Cα distance for contact definition. 15 different distance cutoffs were used, namely, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 10.0, 12.0, 15.0, 20.0, 30.0, 50.0, and 100.0 Å to define contacts. B, HHpred alignments and the Cα-Cα distance were used. C, SUPRB alignments and the Cα-Cα distance were used. D, ROC curves using TM-align alignments and the Cβ-Cβ distance for contact definition. The same 15 different distance thresholds as used for the Cα-Cα distance cutoffs were used. E, HHpred alignments and the Cβ-Cβ distance were used. F, SUPRB alignments and the Cβ-Cβ distance were used. G, ROC curves using TM-align alignments and the heavy atom distance for contact definition. The same 15 different distance thresholds as used for the Cα-Cα distance cutoffs were used. H, HHpred alignments and the heavy atom distance were used. I, SUPRB alignments and the heavy atom distance were used

Yuan et al. BMC Bioinformatics 2012 13:292   doi:10.1186/1471-2105-13-292
Download authors' original image