Re: [modeller_usage] Ligand Introduction from non-homologous protein

[Modeller Caretaker <>]

On 7/16/13 12:15 PM, Rohitesh Gupta wrote:
> I am trying to model a protein using a template that is 85% similar to
> it, but also wants to transfer ligand from another protein that is 10%
> similar.

That should be fine - just give an alignment of the three sequences, 
looking something like

templ1    AAAAAAAA---*
templ2    ------BBBB.*
target    xxxxxxxxxx.*

Ideally you want to take some of the protein from the second template 
(rather than just the ligand) so you get the ligand-protein interactions 
correct. And ideally you want a bit of overlap between the two templates.

> This code runs fine but the ligand doesn't have any of the above
> mentioned restraints.
> How could one instruct modeller to consider these restraints?

Your script looks fine. What makes you think the ligand "doesn't have" 
any of the restraints? Modeller tries to satisfy *all* restraints 
(user-added restraints aren't treated specially) so if there's a 
conflict, you may end up with them not being satisfied.

> I understand that modeller would by default consider the non-bonded
> interaction between ligand and the catalytic site amino acids. In such
> case, is there a way to override those interactions?

Sure, you can add one or more excluded pairs. See
http://salilab.org/modeller/9.12/manual/node108.html
You can just add these in the same special_restraints method.

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