Re: [modeller_usage] Chain designation in completing a structure.

[Modeller Caretaker <>]

On 12/10/09 10:56 AM, Steve Seibold wrote:
> I am attempting to “complete” a structure by placing in a loop and doing
> loop refinement. The problem I am having is setting up the
> “alignment.ali” correctly…I think? The protein has two chains, A and B.
> I am using the “alignment.ali” file shown below: Chain C starts with
> “1”, Chain D starts with “2”, and ends with 1505. So, Chain D has 1504
> residues. The “python script” used for running Modeller is shown below
> the alignment.ali file.
>
> > P1;2PPB_CD
>
> structureX:2PPB_CD: 1 :C:1504:D:undefined:undefined:-1.00:-1.00
...
> return selection(self.residue_range('1244:D', '1250:D'))

When you make a selection, it is using the numbering of the target 
(model) not the template. For consistency, Modeller always numbers the 
target residues starting from 1, and the chains starting at A. So while 
your template may have chains C and D, the model will have chains A and 
B. So you need to modify that selection to use the model numbering - or, 
alternatively, renumber the model residues to your own preference (see 
http://salilab.org/archives/modeller_usage/2009/msg00245.html) Either of 
these changes are only in the Python script though - your alignment file 
looks fine to me.

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