Table 3.
The success rate and the average Z-score of different all-atom potentials compared to that of the DFIRE-SCM potential
| Source | RAPDF-all-atom | KBP-all-atom | DFIRE-all-atom | DFIRE-SCM |
| 4state | 7/7 (3.01)a | 7/7 (3.24) | 6/7 (3.49) | 6/7 (3.94) |
| lattice_ssfit | 8/8 (7.18) | 8/8 (6.61) | 8/8 (9.47) | 8/8 (6.19) |
| lmds | 3/10 (−0.52) | 3/10 (0.53) | 7/10 (0.90) | 3/10 (2.56) |
| fisa | 1/4 (1.27) | 0/4 (1.21) | 3/4 (4.80) | 3/4 (4.70) |
| fisa_casp3 | 3/3 (4.09) | 0/3 (2.08) | 3/3 (5.40) | 3/3 (6.05) |
| CASP4 | 20/23 (2.17) | 20/23 (2.93) | 19/23 (2.61) | 19/23 (3.15) |
| Rosetta | 24/41 (3.18) | 23/41 (3.17) | 31/41 (3.91) | 33/41 (4.90) |
| Summary | 66/96 (2.82 ± 2.87) | 61/96 (3.01 ± 2.46) | 77/96 (3.80 ± 3.31) | 75/96 (4.30 ± 2.22) |
| 75/96 (top5)b | 70/96 (top5) | 84/96 (top5) | 85/96 (top5) |
a The first number is the number of native structures ranked number one; the second number is total number of proteins in the decoy set. The numbers in parentheses are the average Z-scores.
b The first number is the number of native structures that are with top 5 rank.