From Protein Structure to Function with Bioinformatics.pdf

3 Comparative Protein Structure Modelling 79a result, the main chain RMS error rises to about 1.5 Å for about 80% of residues.The rest of the residues are modelled with large errors because the methods generallyfail to model structural distortions and rigid body shifts, and are unable torecover from misalignments. When sequence identity drops below 30%, the mainproblem becomes the identification of related templates and their alignment withthe sequence to be modelled. In general, it can be expected that about 20% of residueswill be misaligned, and consequently incorrectly modelled with an errorlarger than 3 Å, at this level of sequence similarity. These misalignments are aserious impediment for comparative modelling because it appears that most structurallyrelated protein pairs share less than 30% sequence identity (Rost 1999).To put the errors in comparative models into perspective, we list the differencesamong structures of the same protein that have been determined experimentally. A1 Å accuracy of main chain atom positions corresponds to X-ray structures defined ata low-resolution of about 2.5 Å and with an R-factor of about 25% (Ohlendorf 1994),as well as to medium-resolution NMR structures determined from 10 inter-protondistance restraints per residue (Fig. 3.4.). Similarly, differences between the highlyrefined X-ray and NMR structures of the same protein also tend to be about 1 ÅFig. 3.4 Illustrating accuracies of structural models obtained from various experimental andcomputational sources for the same, Der P 2 allergen protein. (A) Superposition of ten alternativeNMR solution structures (PDB code 1A9V) for Der P 2; average RMSD = 0.97 Å. (B)Superposition of X-ray crystallographic structures of two isoforms of Der P 2 protein (2F08(2.20 Å resolution) and 1KTJ (2.15 Å resolution) sharing 87% sequence identity). RMSD =1.33 Å. (C) Superposition of NMR and X-ray solutions of Der P 2 protein (1A9V and 1KTJ).RMSD = 2.2 Å. (D) Superposition of the comparative model built for 1NEP protein using 1KTJas a template and the X-ray solution structure of 1NEP. 1NEP and 1KTJ share 28% sequenceidentity representing a typical difficult comparative modelling scenario. RMSD = 1.66 Å. AllRMSD values refer to Cα superpositions