Hi All,
I would like to create an oligomeric (hexamer) model. I have a Ca- only hexameric model (chains A-F) to use for the template and an experimentally determined monomer structure I would like to use for the target. Essentially, I would like to map six of the monomer target structures to chains A-F of the template. I've attempted to do this using the "model-myini.py" script as described here:
http://www.salilab.org/modeller/manual/node26.html
and here:
http://www.salilab.org/modeller/examples/automodel/model-myini.py
but my output hexamer structure ends up looking like spaghetti. Am I using the proper modeling script for this task? If so, what should the alignment and target PDB files look like to map the monomer structure to the hexamer structure?
Thanks, Doug
As a followup, can I set my experimentally determined monomer structure (target) as six separate rigid bodies (chains A-F), then position the six rigid target structures (chains A-F) separately to a chain of the template (chain A of the template to chain A of the target) based on a sequence alignment, then perform minimization? Is this a good method and can someone explain how I might set this up?
Thanks, Doug
On Aug 6, 2008, at 8:50 PM, Douglas Kojetin wrote:
> Hi All, > > I would like to create an oligomeric (hexamer) model. I have a Ca- > only hexameric model (chains A-F) to use for the template and an > experimentally determined monomer structure I would like to use for > the target. Essentially, I would like to map six of the monomer > target structures to chains A-F of the template. I've attempted to > do this using the "model-myini.py" script as described here: > > http://www.salilab.org/modeller/manual/node26.html > > and here: > > http://www.salilab.org/modeller/examples/automodel/model-myini.py > > but my output hexamer structure ends up looking like spaghetti. Am > I using the proper modeling script for this task? If so, what > should the alignment and target PDB files look like to map the > monomer structure to the hexamer structure? > > Thanks, > Doug
Douglas Kojetin wrote: > I would like to create an oligomeric (hexamer) model. I have a Ca- > only hexameric model (chains A-F) to use for the template and an > experimentally determined monomer structure I would like to use for > the target.
If you have a multimer template, and want to build a multimer model, that's straightforward. Just use the regular modeling scripts, and add chain breaks (/) in your alignment file accordingly. See http://www.salilab.org/modeller/9v4/manual/node28.html for an example. Your alignment would look like
>P1;hextemplate structureX:hextemplate.pdb:1:A:4:F:::: aaaa/bbbb/cccc/dddd/eeee/ffff* >P1;model sequence::::::::: AAAA/BBBB/CCCC/DDDD/EEEE/FFFF*
If you also want to use your monomer structure, just align it with one of the chains of the other template and build a model using both templates. In principle you could use 6 copies of your monomer structure, each aligned with a different chain, e.g.
>P1;hextemplate structureX:hextemplate.pdb:1:A:4:F:::: aaaa/bbbb/cccc/dddd/eeee/ffff* >P1;expmtmonomer1 structureX:expmtmonomer.pdb:1:A:4:A:::: aaaa/----/----/----/----/----* >P1;expmtmonomer2 structureX:expmtmonomer.pdb:1:A:4:A:::: ----/bbbb/----/----/----/----* etc. >P1;model sequence::::::::: AAAA/BBBB/CCCC/DDDD/EEEE/FFFF*
You could also build a monomer model (aligned to just one chain of your template - just change the sequence for the model, leaving the templates unchanged, and replace all but one chain with gaps) but then you have to figure out how to fit your 6 monomer models together after modeling. That sort of protein-protein docking is outside the scope of the Modeller package.
> I've attempted to do > this using the "model-myini.py" script as described here:
No, don't do that. That just sets the starting point for the optimization - not the final structure - and so is only useful in cases where the optimizer needs a 'helping hand' to build the model (for example, you may manually correct serious problems, such as knots in the protein chain, that the optimizer wouldn't be able to fix itself). If you want to influence the *final* structure, include it as a template.
Ben Webb, Modeller Caretaker