[modeller_usage] Re: Building multi-chain models

[Modeller Caretaker <>]

On 6/6/24 4:56 AM, alina.s123567--- via modeller_usage wrote:
> Could you please advise, how I should correctly write this 10 field
> line. I have already read the manual
> https://salilab.org/modeller/manual/node501.html#alignmentformat, but
> there is information about 2 chains. I still can not understand how
> correctly I should write it for 5 chains....
> The alignment for my template is look like:
>> P1;6HUG
> structureX:6HUG:FIRST:A 437:A 473:B 495:C 437:D 473:E:::3.5:-1.00

Modeller is quite simplistic here and can only read a single range of 
residues from the file. It reads the PDB or mmCIF file line by line, 
starting with the first residue you specify and ending with the last 
one. So you can't give it multiple ranges as you have here, i.e.
FIRST:A 437:A 473:B 495:C 437:D 473:E

What might work instead, if the chains are organized alphabetically in 
the file, would be something like

structureX:6HUG:FIRST:A:473:E:::3.5:-1.00

i.e. read starting from the first residue in chain A until residue 473 
in chain E. This won't work though if there are C-terminal or N-terminal 
residues in the intervening chains that you don't want in the model - 
there is no way to tell Modeller not to read them from the PDB file. In 
this case, either include the extra residues in your template sequence 
in the alignment file (and align them with gaps in the model, so they 
don't get used) or edit the PDB file and delete the residues you don't want.

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