Open pgrinaway opened 8 years ago
I don't think we want to do this. What we really need is a way to take a list of molecules
and produce an ffxml
file that defines both templates and any additional parameters.
There is no meaningful way to separate residue templates and the additional parmchk2
-generated parameters at this point.
For me, I add extra atoms and bonds (dummies) after the antechamber and parmchk2 parameters, so I will need separate templates. I can't generate parameters for unphysical atoms, unless I missed an option to mark atoms as dummies to antechamber and openmoltools.
For me, I add extra atoms and bonds (dummies) after the antechamber and parmchk2 parameters, so I will need separate templates. I can't generate parameters for unphysical atoms, unless I missed an option to mark atoms as dummies to antechamber and openmoltools.
You will actually need to generate the ffxml
files yourself since you will also be adding additional information about titration states, relative energies, etc. I don't think we want that to go into openmoltools
for now.
Instead, you will want to use openmoltools
to generate residue templates (and extra parameters) for a set of protonation/tautomer states, and then you will want to convert these to your modified ffxml
format with additional dummy atoms yourself.
I'm confused, how did we previously generate templates? I'm already modifying ffxml files, generated for each state, but I assumed there was a template generator somewhere.
Sorry, the forcefield code is just so confusing. I'm not even sure I need templates if I already have ffxml files.
Let me go through this again with you in person tomorrow.
Sounds good.
As discussed here, it would be useful to either separate parameter and template generation, or store templates for efficiency. Unfortunately, it seems we have partial charges in the template, which complicates this and might require storing the template in the ffxml.