From: Ryan Woltz (rlwoltz_at_ucdavis.edu)
Date: Fri Jun 17 2022 - 17:12:28 CDT

Dear VMD community,

         I've built a few systems for MD using CHARMM-GUI and I'm using VMD
to set restraints by changing beta values. My protein is a tetramer so I'm
selecting by sequence versus resid that way I select everything. Also I
have overlapping residue numbers with my protein and TIP3/Lipids so I
prefer to use sequence. However, when I load a PDB I get this error:

Warning) Unusual bond between residues: 554 (protein) and 555 (none)
Warning) Unusual bond between residues: 555 (none) and 556 (protein)
Warning) Unusual bond between residues: 555 (none) and 556 (protein)

I saw this post and the linked solution from John Stone but the issue here
is different. In there system the protein was forming bonds to ligands or
lipids and therefore could be ignored.

https://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2008-2009/3826.html

    My error is that amino acids are not being recognized as protein. I
can't find any abnormalities in my PDB. If it errors on two different runs
it is not always the same amino acids. In step1_pdbreader.pdb it recognizes
all amino acids. with step5_assembly.pdb it does not recognize 555P, 702L,
1068R. Maybe this has to do with forming artificial bonds with lipids and
thus the amino acids now have unrecognizable structures? The big issue here
is, many of my scripts are selecting segments by sequence. I got around
this the first time I came across this error by breaking up the sequence
that contained the amino acids into 2 flanking sequences. I then went back
into the files and manually altered the offending amino acid. While this
method works it is very prone to me making mistakes as I'm scaling up to
30+ simulations and hundreds of restraint files and I can't really afford
to make a mistake. This error is currently about 5-10% of all systems built.

        Any suggestions on how to force VMD to recognize certain atoms as
part of a protein? I can send any PDBs if needed. I also have psf and parm7
for amber and .crd for namd, can these be used to create restraint pdbs?
When I load the PSF files I get the correct fragments protein 8. When I
load the PDBs I get fragments protein 9, due to the unrecognized amino acid
555 breaking the fragment 3 into multiple sections.

Thank you for any advice you can give,

Ryan