Secondary neutral amine becomes planar due to wrong virtual site construction

GROMACS version: 2022
GROMACS modification: No
Dear dynamicists,
I have parameters for a modified lysine, where the side chain has been extended at the amine, so that the latter is now a neutral secondary amine. pdb2gmx generates a virtual_site3 type 2 parameter for the single hydrogen on the amine, making the amine flat, but it should be tetrahedral. I can change the topology manually and make a virtual_site3 type 4, but that is obviously error prone and cumbersome. Is there a way to avoid this? Shall I report this in GitLab?

Filed an issue https://gitlab.com/gromacs/gromacs/-/issues/4573

Hi,

As far as I understood you have implemented some modifications to a side chain in the force field by changing the rtp file in the force field directory.
If it is the case, did you accordingly modify other force field database files (like Virtual site database) ?
see pdb2gmx input files — GROMACS 2022.2 documentation)

\Alessandra

Hi, not the vsd file in this case, because to my knowledge the vsd files are only used in cases where dummy atoms are needed, like -CH3 groups. And for aromatics, but those aren’t commonly used.

If there is a way to use the vsd files to guide psb2gmx in this case that would be great, but I don’t know how or if that is possible.