From: Paweł Kędzierski (
Date: Mon Oct 17 2022 - 07:11:35 CDT

W dniu 14.10.2022 o 13:45, Neeraj Dhandia pisze:
> Hello VMD users,
> I have come across 2 problems:
> - I am currently trying to make psf file of a protein chain (PDB ID:
> 1USZ), but after running automatic psf builder, it breaks into two chains.
> I am currently using CHARMM V.47 topology files. How could i avoid this?

1UZS contains selenomethionine residues (MSE) which are not recognized
by AutoPSF as protein residues. If your topology files do define such
amino acid and you want it to be preserved (in many cases the seleno
amino acids are not natural, they might have been introduced to help to
resolve the crystal structure), just delete the second chain in AutoPSF
and edit the first one to cover the entire protein (at "Step 3" there
are relevant buttons next to the table of chains).

Otherwise, rename the MSE residues to regular MET in the PDB file before
loading it into VMD; you can do a search & replace in any text editor.

> - Same cystine residue is written twice in the output pdb file with
> different atom positions after running automatic psf builder. I am using
> protein, PDB ID: 3VOR. Should i be worried about this or ignore this??

I can't reproduce this problem. There are 2 different CYS residues in
3VOR connected by disulphide bridge and AutoPSF processes them just fine.



> Any suggestion to understand this behavior would be very helpful.
> Thank you and Best Regards,
> Neeraj Dhandia