ACE ignored when calling addMissingHydrogens

[ross-pure]

I was using this tool to add an ACE cap to my protein and then add missing hydrogens. I noticed that hydrogens were added to all residues except for the ACE. Looking at the code I believe I can see why this happens: when getting the variants for ACE, in the call to _describeVariant, there is the following logic

if residue.name not in app.PDBFile._standardResidues and self._getTemplate(residue.name) is not None:
    # The user has registered a template for this residue.  Use the hydrogens from it.
    ...

This branch is taken because ACE is not in _standardResidues. But the template is just the default one loaded from the templates directory, which doesn't have any hydrogens. As a result, the value returned from _describeVariant is an empty list, instead of None. Consequently, since the Modeller has the following branch:

if isinstance(variant, list):
    spec = Modeller._ResidueData(residue.name)
    infinity = float('Inf')
    spec.hydrogens = [Modeller._Hydrogen(name, parent, infinity, None, None) for name, parent in variant]

no hydrogens will be added.

My questions are:

• Is this expected behaviour?
• What is a reasonable work around to have the hydrogens actually get added to the ACE (and NME) residue(s)?

[peastman]

You're completely right. We only want it to use user-supplied templates there, not ones from the templates directory. That's what residue.name not in app.PDBFile._standardResidues is supposed to accomplish. But really we should record a list of the built in templates and check that instead of _standardResidues.

[peastman]

The fix is in #320, if you want to test it out.

[ross-pure]

Just tested #320 on the system that I was having the issue with and now it seems to be fixed - Hs are added to the ACE. Thanks!

[peastman]

Thanks!