On 6/6/13 4:00 AM, Dermot Mallon wrote: > Between models with > identical sequence the RMSD is ~0.2
This is expected behavior. Modeller models are samples from the set of all models that satisfy the restraints.
> Therefore, I wonder if there is a way to ask modeller to start at the > same point in these molecules? I tried a.rand_model = None, but that > doesn't appear to be the correct option.
It sounds like what you're looking for is rand_method, not rand_model. But Modeller will still optimize to satisfy the restraints, which will move things around. You could just take the initial model (the *.ini file) which is a PDB file of the model before randomization and optimization. This won't change unless you change your alignment and/or template structures. But since it's unoptimized it could contain clashes.
Ben Webb, Modeller Caretaker