From: Maxim Belkin (
Date: Fri Feb 05 2016 - 14:20:04 CST

That’s right — you need to build a segment first. Couple of comments: you probably don’t need "guesscoord" as you don’t have any missing atoms. Also, your segment name is too long. Try loading the newly created new.p{sf,db} and do the following in the tk console:

set sel [atomselect top all]
puts   "[lsort -unique [$sel get segname]]"

And see what segment name you see.


On Feb 4, 2016, at 21:23, Olya Kravchenko <> wrote:

Sorry for cluttering, I've found an error. In my psfgen script I
missed a segment line. Now it looks like this:

package require psfgen
topology my_topology.inp
segment argon {pdb argon.pdb}
coordpdb argon.pdb argon
writepdb new.pdb
writepsf new.psf

This way everything works.


On Thu, Feb 4, 2016 at 9:53 PM, Olya Kravchenko <> wrote:
I still have a problem, now with generating psf file. I'd very much
appreciate your input.

I would like to create a psf file for my simulation box that consists
of 200 argon atoms. I have generated a pdb file for argon, here is a
portion of it:

ATOM      1  AR  AR  A   1      18.866  14.256   9.507  0.00  0.00
ATOM      2  AR  AR  A   2       6.942  41.521  12.453  0.00  0.00
ATOM      3  AR  AR  A   3      48.095  26.528   6.211  0.00  0.00

My topology file for argon (found in the archives where similar issue
was discussed):

*>>>>>> CHARMM topology for Argon <<<<<<<<<<
27  1
MASS     1 AR      39.95 AR ! ARGON
RESI AR           0.00   !Argon
ATOM AR  AR       0.00

Here is what I use to create psf:

package require psfgen
topology my_topology.inp
coordpdb argon.pdb
writepdb new.pdb
writepsf new.psf

After I type coordpdb argon.pdb in tk console I get warnings for all
the atoms in my pdb file:

coordpdb test.pdb
psfgen) reading coordinates from pdb file test.pdb
psfgen) no segment
psfgen) Warning: failed to set coordinate for atom AR AR:1
psfgen) no segment
psfgen) Warning: failed to set coordinate for atom AR AR:2
psfgen) no segment
psfgen) Warning: failed to set coordinate for atom AR AR:3
psfgen) no segment

Then guesscoord gives me:

psfgen) Info: guessing coordinates for 0 atoms (0 non-hydrogen)

Then empty psf and pdb files are created.

The atom names in pdb file and topology file are consistent. What am I
missing? It must be something simple that I overlook.

Thank you!!


On Sat, Jan 30, 2016 at 8:28 PM, Peter Freddolino <> wrote:
(The solvate docs also have info on this:
That page has been missed in some previous threads)
On Jan 29, 2016, at 3:18 PM, Maxim Belkin <> wrote:

Sorry, my mistake:

solvate -minmax {{-50 -50 -50} {50 50 50}} -o water_box

For non-standard solvent you can use solvate too. Just type solvate and it should give you help on that. VMD mailing list should have the info you are looking for.


On Jan 29, 2016, at 13:36, Olya Kravchenko <> wrote:

Thanks a lot!!

After running this command

solvate -minmax {{-50 50 -50} {50 50 50}} -o water_box

I am getting the following error:

Couldn't load output files. Please make sure that you have enough disk
space, and that you didn't set solvent box boundaries that would
produce zero placed waters.

How should I address it?

Also, is there a way to create a box filled with atoms/molecules other
than water?

On Fri, Jan 29, 2016 at 2:26 PM, Maxim Belkin <> wrote:
package require solvate
solvate -minmax {{-50 -50 -50} {50 50 50}} -o water_box

On Jan 29, 2016, at 13:17, Olya Kravchenko <> wrote:

Hi all,

I would like to create a simple rectangular simulation box filled with
spherical particles (such as Argon atoms) for testing purposes. What
would be the easiest way to create such system? Is it even possible
with vmd or namd?

At this point, I can only think of using a solvate plugin and creating
a box around small protein, then deleting the protein and
equilibrating the box without it, but this doesn't seem optimal. Is
there another plugin/method that would be more suitable?

Thanks in advance!