Chinwe Ekenna wrote: > i have successfully created protein structures using modeller 9v4 but > i am facing a new challenge ..... the structures are to be further > used in creating a docking site for identified drug targets..... the > challenge i'm facing is that all my structure result as regards the > docking site say no site found(using Qsitefinder).... this sir is not > possible as they should have docking sites.. what do you advise that > i do.....
Modeller will build comparative models, thus by construction they will look very similar to the templates. If there are no suitable binding sites in the templates, or your sequence identity is very low in these regions, it is likely that the model will not have any suitable binding sites either.
One thing to note is that if your template is already bound to a ligand, it may make sense to include that ligand in your model. If nothing else, it will ensure that the binding site geometry is preserved (you can always remove the ligand later). If on the other hand you are trying to construct a binding site where none exists in the template, that is not likely to work. You would have to take your models and relax them with an MD package, or possibly add some extra restraints to "force" the modeling to do what you want.
Ben Webb, Modeller Caretaker