Root-mean-square-deviation Modeller comparisons

Figure 2-4. Comparison of optimized and X-ray structures for the active site of RNR. The X-ray structure of R2met is superimposed on the optimized structures from active-site QM-only (left) and ONIOM2 0middle) models. The plot shows the quality of the optimizations evaluated as the root-mean-square deviations (in A) compared to the X-ray structures of RNR and MMO (right). (Adapted from Torrent et al. [24]. Reprinted with permission. Copyright 2002 Wiley Periodicals, Inc.)...

It is quite instructive to compare these new measurements (which lie outside the data bases available at the time the various mass models were formulated) with predictions from the models. For such comparisons it is convenient to define A = Predicted Mass - Measured Mass. A > 0 thus denotes cases where the binding energy has been predicted to be too low and conversely, A < 0 corresponds to a prediction of too much nuclear binding. Table 1 summarizes average and root-mean-square deviations for twelve models. 

The comparisons between the experimental and calculated composition of each component of two phases were made through root mean square deviation (rmsd). The rmsd were calculated from the difference between the experimental data and predictions of the GMDH model at different temperatures according to the following equation ... 

Figure 15.6. Molecular model comparison of wild-type and amino terminal fragment mutant pepsin. Superposition of the bottom -sheet and N-terminal fragment of wUd-type and N-frag mutant models. There were no major differences. Root-mean-square deviation between backbone atoms of wUd-type and N-frag was 0.28 A. The numbaa show the mutation sites in N-frag.

Comparisons between optimized and X-ray structures were once again made by calculating root-mean-square (RMS) deviations. When comparing all heavy atoms in the protein, the total RMS deviation is approximately 1.7 A, irrespective of method for the model system or the ONIOM implementation (mechanical, ONIOM-ME, or electronic embedding, ONIOM-EE). The largest deviations occur for residues in the vicinity of the second monomer. Therefore, adding the second monomer to the model should improve the calculated geometries. 

---