Anna Marabotti wrote: > I have a question about the possibility to introduce a restraint into > a model. I obtained a model of a protein, and looking at the > structure I saw that an Arg residue lies in proximity of an Asp > residue, but their charged moieties are not positioned properly to > infer a ion pair interaction between them. Since from other > experimental data I strongly suspect that this ion pair is present > (but I am not 100% certain), I would like to know if I can force > somehow these two residues to interact each others, without > disrupting the rest of the model. The problem is not only on the side > chain, because I cannot simply rotate it or use some programs like > SCWRL or so on to change the position of the side chains: also the > backbone would be affected if such an ion pair would be present. > Therefore I would like to know if I can add some "soft" restraints, > for example to "ask" MODELLER: keep the charged moieties of Arg and > Asp not farther than X A only if the backbone angles are changing not > more than Y degrees to do it. Is there a way to do it?
No, there is no way to do this directly with the default set of restraints in Modeller. So you have two options:
1. Add a distance restraint between the two residues. Build models with this distance restraint and compare with models built without the distance restraint, to see if the resulting change in backbone angles fits your criteria.
2. Add your own user-defined restraint. See http://salilab.org/modeller/9v6/manual/node423.html. For example, you could write a restraint that works like a distance restraint, but only if some set of angles currently lie within your criteria (otherwise the energy is reported as zero). This is, however, more complicated to achieve.
Ben Webb, Modeller Caretaker