> -----Original Message----- > From: Modeller Caretaker [mailto:modeller-care@salilab.org] > Sent: Wednesday, December 07, 2016 9:00 AM > To: Markus Heller; modeller_usage@salilab.org > Subject: Re: [modeller_usage] Modeling loop using template from PDB > file > > On 12/6/16 3:54 PM, Markus Heller wrote: > > I have a structure (A) which is missing a stretch of a dozen or so > > residues. I also have the structure of the protein from a different > > organism which has said stretch of residues resolved (B). Now I'd > > like to take the coordinates of A and model the missing residues > using > > the relevant coordinates (i.e. for these residues only) from B. > > > > How would I go about doing this? > > Just make an alignment containing both proteins A and B, and align the > missing region with B and the rest with A, as per > https://salilab.org/modeller/FAQ.html#1 > > If you're happy with the A coordinates and don't want them to move at > all, select only the B region in select_atoms(): > https://salilab.org/modeller/9.17/manual/node23.html

I have it working with modeling for 1 chain with the following class definition:

class MyModel(automodel): def select_atoms(self): # Select residues 1 and 2 (PDB numbering) return selection(self.residue_range('339:A', '372:A'))

I can't get it to work for both chains.

This

class MyModel(automodel): def select_atoms(self): # Select residues 1 and 2 (PDB numbering) return selection(self.residue_range('339:A', '372:A', '809:B', '842:B'))

gives:

TypeError: residue_range() takes exactly 3 arguments (5 given)

Each chain is 470 AA, so 339:A and 372:A should be 809:B and 842:B in modeller nnumbering.

On the other hand, this, assuming we're using PDB numbering here, according to which 339 through 372 exist in both chains,

class MyModel(automodel): def select_atoms(self): # Select residues 1 and 2 (PDB numbering) return selection(self.residue_range('339:', '372:'))

gives

KeyError: 'No such residue: 339:'

What's my mistake?

Thanks Markus