[modeller_usage] Re: DNA atoms changed by modeller

[Modeller Caretaker <>]

On 11/6/23 7:31 AM, Pedro Guillem via modeller_usage wrote:
> I have a PDB with a small protein-DNA complex... I induce a point 
> mutation in the protein and then save the mutated PDB. The DNA molecule 
> remains untouched.
> 
> I noticed my original pdb defines the first base (adenine) with 18 atoms 
> - but the output modeller builds 21 atoms. The first 3 atoms are added.
> 
> I believe the original input is missing the phosphate group. But I want 
> modeller to leave the DNA alone the way it is. Is there a way to set 
> modeller to leave some chains untouched in the output?

It sounds like you might be using the script at
https://salilab.org/modeller/wiki/Mutate_model ?

That script creates the output model by first determining its primary 
sequence (by copying the original template sequence with append_model() 
then mutating it with mutate()), then constructing the internal 
representation from the CHARMM topology library (generate_topology()), 
copying coordinates from the template (generally for the unmutated part 
of the structure) using transfer_xyz(), then finally filling in any 
unknown coordinates (generally the mutated residue's sidechain) using 
internal coordinates (build()).

This usually works fine for unmodified amino acid chains, as 
generate_topology() will automatically patch the C- and N-termini where 
there is no peptide bond, but no automatic patching is done for nucleic 
acids. In this case it sounds like you want to apply the 5TER patch to 
your first DNA residue using Model.patch() after generate_topology():
https://salilab.org/modeller/10.4/manual/node188.html

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