Hi all
We are trying to build a homotetramer model but having some difficulty maintaining symmetry. I've read the previous posts but maybe someone can help me further - I'm new to Modeller and to Python.
I have imposed symmetry:
s1 = selection(self.chains['A']).only_atom_types('CA') s2 = selection(self.chains['B']).only_atom_types('CA') s3 = selection(self.chains['C']).only_atom_types('CA') s4 = selection(self.chains['D']).only_atom_types('CA') self.restraints.symmetry.append(symmetry(s1, s2, 1.0)) self.restraints.symmetry.append(symmetry(s1, s3, 1.0)) self.restraints.symmetry.append(symmetry(s1, s4, 1.0))
but when I run automodel the molecular dynamics step pulls the monomers away from the template - so that although the subunits are identical, the symmetry is broken.
We can think of a few ways to get around this: 1) do no molecular dynamics, only minimisation (maybe someone can tell me how?) 2) run molecular dynamics only on selected regions of the model so that the core remains fixed in place (again, how would I do this?) 3) learn Python and write restraints using symmetry matrices (if someone has implemented something like this before I'd love to see their code!)
Looking forward to your input Jean
Jean Watermeyer Division of Medical Biochemistry IIDMM University of Cape Town Health Sciences Anzio Rd Observatory 7925 South Africa
Tel. 021 4066832
JEAN WATERMEYER wrote: > We are trying to build a homotetramer model but having some difficulty > maintaining symmetry. I've read the previous posts but maybe someone > can help me further - I'm new to Modeller and to Python.
It's hard to tell from the code snippet. It looks correct (I'm assuming this code is from the special_restraints method?) but maybe you're not using the class properly elsewhere in the script - can you post your complete script? The most common mistake is to derive a 'mymodel' class from 'automodel' but then to instantiate a regular 'automodel' object rather than the new 'mymodel', in which case any additional restraints, patches etc. in 'mymodel' are not applied.
> We can think of a few ways to get around this: > 1) do no molecular dynamics, only minimisation (maybe someone can tell > me how?)
Sure, you can set automodel.md_level to None, which will turn off the MD refinement part. But if MD is not maintaining the symmetry, it seems unlikely to me that CG minimization would. It is more likely that there's a problem with your energy function.
> 2) run molecular dynamics only on selected regions of the model so that > the core remains fixed in place (again, how would I do this?)
Doesn't http://salilab.org/modeller/manual/node23.html do what you need here?
> 3) learn Python and write restraints using symmetry matrices (if someone > has implemented something like this before I'd love to see their code!)
I personally haven't come across such a restraint, but if anybody has written one and wishes to contribute it, I'd recommend the Modeller wiki for this purpose.
Ben Webb, Modeller Caretaker
-
JEAN WATERMEYER -
Modeller Caretaker