Re: [modeller_usage] query about chimeric protein

[Modeller Caretaker <modeller-care@ucsf.edu>]

Sana Bilal wrote:
> im trying to build a chimeric protein based on three known structures 
> but the resultant model is not folded. How can i orient the template 
> structures before modeling?

It sounds like you have not told Modeller what the restraints between 
your three domains should be. See 
http://salilab.org/modeller/9v1/manual/node27.html for some discussion 
on this.

You can orient the template structures in any way you choose - Modeller 
has an initial_malign3d option for this, but it would probably not be 
useful to you unless your three structures had significant structural 
overlap, and it sounds like that isn't the case here. But that would 
only affect your initial model anyway - for the optimization Modeller 
deals purely with distance restraints, not absolute position restraints, 
and each distance comes from a single template, so it is not affected by 
the orientation of different templates. If you can include templates 
which span the interfaces between your three domains, or add manual 
restraints on distances or angles (e.g. from other experimental data) 
then you are much more likely to get a good model.

	Ben Webb, Modeller Caretaker
--
modeller-care@ucsf.edu             http://www.salilab.org/modeller/
Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage