Table 3.
Statistics for PASD-generated Structuresa
| System | CVN | IL-4b | YmoA | mth1743c | NiRD | ThTP overall | ThTP dom. 1 | ThTP dom. 2 | ThTP dom. 3 | EIN |
|---|---|---|---|---|---|---|---|---|---|---|
| Reference Structure | 2EZM | 1ITI | 2JVP | 1RYJ | 2JO6 | 2JMU | 2JMU | 2JMU | 2JMU | 2EZA |
| method | NOE+RDC | NOE | NOE+RDC | NOE | NOE | NOE | NOE | NOE | NOE | NOE+T1/T2 |
| defined residues | 1-101 | 8-129 | 2-67 | 7-69 | 3-108 | 5-215 | 10-34,122-188 | 38-117 | 194-215 | 1-259 |
| backbone precision (Å) | 0.2±0.02 | 0.4±0.03 | 0.4±0.1 | 0.7±0.1 | 0.8±0.1 | 1.1±0.2 | 1.0±0.2 | 0.8±0.2 | 0.7±0.1 | 0.8±0.2 |
| heavy atom precision (Å) | 0.5±0.03 | 0.8±0.03 | 0.5±0.1 | 1.3±0.1 | 1.4±0.1 | 1.6±0.2 | 1.6±0.2 | 1.3±0.2 | 1.2±0.2 | 1.1±0.2 |
| PASD Structure Results | ||||||||||
| backbone accuracy (Å) | 0.9 | 1.3 | 1.9 | 1.6 | 1.2 | 13.3 | 2.5 | 2.7 | 1.6 | 16.4 |
| heavy atom accuracy (Å) | 1.4 | 2.1 | 3.0 | 2.2 | 2.0 | 13.3 | 3.2 | 3.2 | 2.1 | 16.9 |
| backbone precision (Å) | 0.7±0.2 | 0.7±0.1 | 0.8±0.2 | 1.2±0.5 | 1.1±0.3 | 5.9±3.3 | 2.9±1.0 | 2.7±0.9 | 1.3±0.7 | 13.4±1.8 |
| heavy atom precision (Å) | 1.2±0.2 | 1.3±0.1 | 1.4±0.2 | 1.7±0.5 | 1.6±0.4 | 6.4±3.3 | 3.6±1.1 | 3.3±0.9 | 2.0±0.8 | 14.0±1.8 |
| PASD NOE Assignment Statistics | ||||||||||
| coverage | 4.7 | 2.5 | 4.6 | 2.2 | 8.7 | 3.8 | 5.5 | 7.7 | 0.9 | 0.06 |
| discrimination (%) | 86.2 | 81.8 | 91.1 | 76.3 | 88.3 | 64.8 | 69.5 | 69.0 | 77.2 | 67.8 |
For each dataset, the reference structure PDB ID, the published structurally-defined region, and the published coordinate precision are listed, together with the coordinate precision and accuracy of the converged structures from the PASD calculation. The coordinate accuracy is defined as the r.m.s. difference between the regularized mean coordinates of the 50 converged structures and the reference structure’s coordinates, calculated over the published set of structurally-defined residues. The coordinate precision is defined as the mean r.m.s. difference between the 50 converged structures and their regularized mean coordinates. In addition, three measures of NOE assignment quality are reported: NOE coverage, that is the per-residue number of high-likelihood (> 90%) peak assignments with long-range (> 5 residues apart in primary sequence) shift assignments; and NOE discrimination, that is the percentage of peak assignments with likelihood > 90% or < 10%. Residual dipolar coupling restraints were included in the reference structure calculation for CVN and YmoA, but not for IL-4, mth1743, NirD, or ThTP.
Comparison with the IL4 X-ray structure (PDB ID 1RCB) solved at 2.25 Å resolution yields very similar results: 1.3 Å and 2.3 Å backbone and heavy atom accuracy, respectively.
Because the published structure of mth1743 was calculated without torsion angle restraints of any kind, we refined it, starting from the published coordinates 1RYJ, using the published assigned NOE restraints (PDB ID 1RYJ.mr) and TALOS restraints determined from its published chemical shift table (BMRB accession code 5106, Yee et al., 2002), using a standard simulated annealing refinement protocol in Xplor-NIH (Schwieters et al., 2006) which included the same non-PASD energy terms as used in the PASD structure calculations.