How similar must a template protein be for homology modeling by side-chain packing methods?

摘要: Given the correct backbone coordinates of a globular protein, side-chain packing methods can be generally expected to predict buried core residues accurately. In context study in modeling family bacteriophage DNA-binding proteins, we observed that when actual perfect are not available, still predictive value using homologous but imperfect backbones. This is situation practical homology where target protein sequence modeled from template structures known homologs. order assess quality and degree accuracy such predictions their dependence on extent homology, have now extended these studies well characterized globin span much wider range sequence-structure similarity. The collective results show clear relationship independent between prediction level similarity proteins. We judge this terms identity r.m.s. deviation structure used for cases available. summary, as drops 100% about 50%, or increases 0 A 1 A, overall average error buried-core rises 1.2 1.5 while chi 85% 70-75% 2 80% 60-65%. When below 50% above all 3 measures decrease rapidly. edges so-called twilight zone at around 22%, exceeds approaches values random predictions, namely, 3.1 error, 22% 29% prediction. These observations provide evaluation homology-modeler. which topology fold constrain internal orientation gives insight into plasticity found architecture