Don't alter termini

[Dan-Burns]

I glycosylated a structure that was prepared for MD with proper terminal hydrogens but the output PDB has the terminal hydrogens removed.

Any way to retain these?

Thank you!

[Lachele]

In the file I just downloaded, there are no hydrogens at all.

Were the hydrogens specially placed or do you just want them to be able to indicate the charges (or lack there of) on the N and C terminals?

How are you wanting to use this file? Are you planning to use this with AMBER? There are options you can use right now, but it will take some time to put that capability into the builder. We have essentially no funding and minimal staff.

[gitoliver]

Sounds like a feature request. Can you attach your input file and specify which hydrogens you want retained?

[Dan-Burns]

In the file I just downloaded, there are no hydrogens at all.

Were the hydrogens specially placed or do you just want them to be able to indicate the charges (or lack there of) on the N and C terminals?

How are you wanting to use this file? Are you planning to use this with AMBER? There are options you can use right now, but it will take some time to put that capability into the builder. We have essentially no funding and minimal staff.

I'd like to simulate in OpenMM.

I have to use pdbfixer on the glycosylated structure to add hydrogens back in but that doesn't resolve my problem with creating an OpenMM system.

The glycam forcefield is available but when I try to generate a system, it chokes on the first residue bonded to a glycosylated residue.

[gitoliver]

I've not used OpenMM so you'll have to specify what you need our output to look like.

[Lachele]

I don't know OpenMM either, but the general strategy for building is that the hydrogens are added back in via templates containing geometrical information. Amber does that during the process of creating MD input files.

Can you upload an example PDB file that is ready for use in creating the OpenMM system?

[Dan-Burns]

nicastrin.pdb can be simulated in openmm and has 16 sites that can be glycosylated.
nicastrin_glycam.pdb is glycosylated.

nicastrin.pdb.tar.gz

nicastrin_glycam.pdb.tar.gz

You can conda install the latest openMM and then run

from openmm import *
from openmm.app import *

pdb = PDBFile('nicastrin_glycam.pdb')
forcefield = ForceField('amber14-all.xml', 'amber14/tip3p.xml', 'amber14/GLYCAM_06j-1.xml')
modeller = Modeller(pdb.topology, pdb.positions)
modeller.addHydrogens(forcefield=forcefield)

The above will return the following error.

ValueError: No template found for residue 10 (LEU).  The set of atoms matches LEU, but the bonds are different.  For more information, see https://github.com/openmm/openmm/wiki/Frequently-Asked-Questions#template

LEU 10 is connected to an NLN residue which is defined in amber14/GLYCAM_06j-1.xml.

This may need to be raised on the OpenMM issues.

Thank you.

[Dan-Burns]

I believe the problem is that some of the NLN ND2 atoms have 2 hydrogens (HD21 & HD22). For instance, NLN 12 ND2 has 2 and NLN 384 has one.

If I delete the HD22 on NLN 12, OpenMM will fail on the next NLN with more than 2 ND2 hydrogens.

You may need to generate a PDB with conect records for the entire structure to get this result.

[Dan-Burns]

The ideal output from Glycam would retain all the original atoms (except when converting ASN to NLN where HD21 would be removed) and names and conect records and only add the glycosylated residues with conect records for the modified residues, glycans, and residues directly connected to the modified residues.