Table 3. Benchmark results of SCOP superfamily predictions by DBD-Threader.
| SCOP ID | True | C/U/M/I | <TM-score> | <MCC> | SCOP Superfamily Description |
| 46689 | 22 | 16/2/2/2 | 0.69 | 0.57 | Homeodomain-like |
| 46785 | 14 | 11/2/1/0 | 0.65 | 0.38 | Winged helix DBD |
| 47113 | 6 | 6/0/0/0 | 0.65 | 0.56 | Histone-fold |
| 57959 | 5 | 2/2/0/1 | 0.55 | 0.75 | Leucine zipper |
| 47459 | 5 | 5/0/0/0 | 0.79 | 0.78 | HLH DBD |
| 47413 | 5 | 4/0/0/1 | 0.51 | 0.56 | λ repressor-like DBD |
| 56672 | 5 | 5/0/0/0 | 0.71 | 0.36 | DNA/RNA polymerases |
| 52980 | 5 | 3/2/0/0 | 0.53 | 0.25 | Restriction endonuclease-like |
| 47954 | 4 | 4/0/0/0 | 0.86 | 0.56 | Cyclin-like |
| 46894 | 3 | 2/0/0/1 | 0.69 | 0.55 | C-term Domain of bipartite response regulators |
| 56349 | 3 | 1/2/0/0 | 0.60 | 0.50 | DNA breaking-rejoining enzymes |
| 55455 | 2 | 2/0/0/0 | 0.67 | 0.79 | SRF-like |
| 47095 | 2 | 2/0/0/0 | 0.74 | 0.55 | HMG-box |
| 52141 | 2 | 1/1/0/0 | 0.56 | 0.36 | Uracil-DNA glycosylase-like |
| 100939 | 2 | 2/0/0/0 | 0.70 | 0.41 | SPOC domain-like |
| 57701 | 2 | 0/2/0/0 | 0.43 | 0.59 | Zn2/Cys6 DBD |
| 55608 | 2 | 0/1/1/0 | 0.61 | 0.49 | Homing endonucleases |
| 57667 | 2 | 2/0/0/0 | 0.68 | 0.59 | C2H2 and C2HC zinc fingers |
| 46596 | 1 | 0/0/1/0 | − | − | DNA topoisomerase I, dispensable insert domain |
| 56219 | 1 | 1/0/0/0 | 0.69 | 0.45 | DNase I-like |
| 49417 | 1 | 1/0/0/0 | 0.71 | 0.61 | p53-like transcription factors |
| 81624 | 1 | 1/0/0/0 | 0.78 | 0.75 | N-term domain of MutM-like |
| 46946 | 1 | 1/0/0/0 | 0.72 | 0.54 | S13-like H2TH domain |
| 56741 | 1 | 0/0/1/0 | − | − | Eukaryotic DNA topoisomerase I, N-term DBD |
| 48150 | 1 | 1/0/0/0 | 0.79 | 0.85 | DNA-glycosylase |
| 81585 | 1 | 0/1/0/0 | 0.82 | 0.82 | DNA polymerase β-like, second domain |
| 47802 | 1 | 0/0/0/1 | 0.48 | 0.39 | DNA polymerase β, N-term domain-like |
| 48295 | 1 | 0/0/0/1 | 0.27 | 0.17 | TrpR-like |
| 57716 | 1 | 0/0/1/0 | − | − | Glucocorticoid receptor-like (DBD) |
| 81301 | 1 | 0/0/1/0 | − | − | Nucleotidyltransferase |
| 53335 | 1 | 0/1/0/0 | 0.83 | 0.61 | SAM-dependent methyltransferases |
| 100879 | 1 | 0/1/0/0 | 0.80 | 0.79 | Lesion bypass DNA polymerase, little finger domain |
| 48334 | 1 | 0/1/0/0 | 0.82 | 0.68 | DNA repair protein MutS, domain III |
| 47823 | 1 | 0/1/0/0 | 0.68 | 0.52 | λ integrase-like, N-term domain |
| 47781 | 1 | 0/1/0/0 | 0.30 | 0.27 | RuvA domain 2-like |
| 55811 | 1 | 0/0/1/0 | − | − | Nudix |
| 53098 | 1 | 0/0/1/0 | − | − | Ribonuclease H-like |
| Total | 110 | 73/20/10/7 |
The results are based on 110 DNA-binding domains from 84 predicted DNA-binding proteins that have official (True) SCOP assignments. The numbers of SCOP domain predictions are shown for four groups: Consistent (C) Un-annotated (U), Inconsistent (I), and Missed (M). Consistent predictions have the same SCOP superfamily assignments as the true assignments. If a DNA-binding domain is predicted based on a template without an official SCOP assignment, then the SCOP prediction is skipped and the domain belongs to the Un-annotated group. A target protein may contain multiple DNA-binding domains, and any domain not detected by DBD-Threader is a Missed domain. The remaining SCOP predictions, in which SCOP predictions are different from the official assignments, are considered Inconsistent. < > denotes mean. TM-scores and MCCs of DNA-binding residue predictions were calculated for all predicted domains per SCOP superfamily.