Table 4.
Percentage of good Best-of-All alignments missed by each method
| Relaxed testa | Best-of-All testa | Test set size | SSAP | STRUCTAL | DALI | LSQMAN | CE | SSM |
|---|---|---|---|---|---|---|---|---|
| SAS<6 Å, Nmat>40, overlap>50% | SAS<3 Å, Nmat>50, overlap>60% | 20,065 | 16.4 | 1.5 | 29.6 | 55 | 4.3 | 3.9 |
| SAS<6 Å, Nmat>35, overlap>0% | SAS<3 Å, Nmat>45, overlap>0% | 33,485 | 32.4 | 0.3 | 12.3 | 8.6 | 6.1 | 1.5 |
| SAS<7 Å, Nmat>40, overlap>50% | SAS<4 Å, Nmat>50, overlap>60% | 66,125 | 29.4 | 4.4 | 22.4 | 74.5 | 7.5 | 9.9 |
| SAS<8 Å, Nmat>40, overlap>50% | SAS<5 Å, Nmat>50, overlap>60% | 138,373 | 41.2 | 7.2 | 22.9 | 85.1 | 11 | 17.6 |
| SAS<7 Å, Nmat>35, overlapa>0% | SAS<4 Å, Nmat>45, overlap>0% | 140,826 | 51.8 | 0.2 | 12.1 | 20.5 | 7.7 | 1.8 |
| SAS<8 Å, Nmat>35, overlap>0% | SAS<5 Å, Nmat>45, overlap>60% | 349,203 | 64.3 | 0.3 | 20 | 44.3 | 10.8 | 3.1 |
A good Best-of-All has a SAS score less than SAS defined in the Best-of-All test, and an alignment missed by a particular method has a SAS score greater than SAS defined in the relaxed test.
a
Nmat is the size of the alignment. Overlap is defined as the percentage of the shorter structure that is matched.