From: Josh Vermaas (
Date: Mon Apr 08 2013 - 18:18:59 CDT

Hi Zachary,

While I can't speak to why the GUI version of mergemols crashes (I can
only guess the GUI of mergemols actually tries to recreate a new psf
using psfgen for your small molecule, and then fails miserably somehow),
I can offer a possible solution! TopoTools is a plugin to VMD that
merges molecules in a way I've found to be more robust than the GUI, and
is quite well documented: You'll be most
interested in the mergemols command (near the bottom of the page). You
should inspect your small molecule psf very carefully though before
using it, to make sure it has all the angle/dihedral terms you'd expect,
as I've had cases where webservers will make a psf without any angle or
dihedral terms, and TopoTools will not check that.

-Josh Vermaas

On 04/08/2013 04:54 PM, Zachary E Tibbs wrote:
> Hello all,
> I would eventually like to runa NAMD/CHARMM simulation on my
> system.But, I am running intoa problemwhen using VMD to prep the
> system--The problem occurs while trying to merge two pdb/psf pairs.
> One of these files is a protein for which I generated a psf for
> through psfgen, while the other is a small moleculewhich I submitted
> to the SwissParam server for psf generation/parameter calculations. I
> can open each pair in VMD using "mol new filename.psf; mol addfile
> filename.pdb"with no issues, leading me to believe that all files are
> in good order. I attempted to use the 'Merge Structures' extension to
> merge the two structures, as I have in the past. An update pops up
> stating, "No more segment name conflicts were detected...Merging
> structures,"as normal.After clicking OK, I receive an error message
> "MOLECULE DESTROYED BY FATAL ERROR! ..." The details of the error lead
> me to believe that the error occurs while reading the small molecule
> psf--The error message is posted below.
> while executing
> "readpsf $m2psf"
> invoked from within
> "psfcontext eval $psfcon {
> # this is a hack, just to get the topology file
> package require readcharmmtop
> set topologyfile [format "%s/top_a..."
> (procedure "mergeMolecules" line 6)
> invoked from within
> "mergeMolecules [list "$guiState(mergef1src).psf"
> "$guiState(mergef1src).pdb" ] [list "$guiState(mergef2src).psf"
> "$guiState(mergef2src).pdb" ] $gu..."
> (procedure "::MergeStructs::guiMergeConflictsResolved" line 6)
> invoked from within
> "::MergeStructs::guiMergeConflictsResolved"
> ("after" script)
> I immediately attributed the problem to the fact that the small
> moleculepsf was generated by the above named server--Somethingwhich I
> am new to use.But, after manual inspection of the psffiles side by
> side, Icame to the conclusion that the small molecule psf has a nearly
> identical setup as the protein psf--So I amat a loss for identifying
> the problem. It seems there is an issue with the small molecule psf
> file....But I assumedthat aproblem with the psf would disallow VMD to
> even read the file in the first place to open the structure.Has anyone
> experienced a similar issue who could offer guidance to resolve this
> problem? Any help would be tremendously appreciated. Thank you.
> Z