Using VMD and NAMD on MOGON

We, the ZDV HPC group, have setup and compiled VMD and NAMD for you – and we will continue to do so for new versions upon request. You will find the installed modules as

vis/VMD/<version_string>
chem/NAMD/<version_string>

Here, the <version_string> may deviate between the clusters. If you are missing a particular version, please inform us.

That being written, we are aware that starting the VMD-GUI can be tricky and the relevant information hard to find.

Here, we give some brief snippets, covering the essentials:

• First, we need to load the module and instruct VMD to start in its graphical, non-MPI mode:

$ module load vis/VMD # this will load the most current version, which is installed
$ export VMDNOMPI=1

• Next, we start an interactive session, e.g. with salloc. Here is a little more information on interactive jobs. With respect to the number of cpus: Not everything in VMD is parallelized. If you want to start an VMD-MPI job, you better write a VMD script and start VMD like a conventional MPI-application.

$ salloc -A <your_account> -p smp -c <number_of_cpus> -t <sufficient_time>
<snipping wainting for the session>
$ srun vmd <arguments>

• When ending the session, do not forget to type exit to relinquish the allocation.

A .psf (protein structure file) ist needed to feed NAMD. Beginners frequently have troubles generating one.

The following section refers to the VMD tutorial section on generating a PSF file:

1. load a new molecule
2. run the autopsf utility (Extensions → Modeling → Automatic PSF Builder)

This will create the following files: <molecule_prefix>_autopsf_formatted.pdb, molecule_prefix>_autopsf.log, <molecule_prefix>_autopsf.pdb, <molecule_prefix>_autopsf.psf.

Of course, VMD offers many more options, most are more sophisticated than this description.

To run a simulation with periodic boundaries, we assume that you have loaded the desired .pdb file into VMD and will set to calculate the unit cell:

Within VMDs Tk Console²⁾ you can run:

set everyone [atomselect top all]
measure minmax $everyone 

This will yield two vectors, e.g.: {-38.473999 -24.975999 -35.2630} {36.910999 29.686 35.957}. To obtain the values for the initial(!) cellBasisVector-entries you need to add together the absolute values, e.g. abs(-38.473999) + 36.910999 = 75.384998. Subsequently set the cellBasisVector-entries to sensible (slightly rounded) values, such that the result vector of your addition becomes the diagonal:

cellBasisVector1     75.385 0.0 0.0
cellBasisVector2     0.0 54.662 0.0
cellBasisVector3     0.0 0.0 71.22
cellOrigin           0.0 0.0 0.0

Finally, turn on PME-settings and set the box size, such that the initial conformation will surely fit, e.g.:

PME                  on
PMEGridSizeX         80
PMEGridSizeY         60
PMEGridSizeZ         75

With a given configuration file, you can start NAMD like this:

#!/bin/bash
 
#SBATCH -t <time>
#SBATCH -p <partition> # e.g. nodeshort / parallel
#SBATCH -A <account>
#SBATCH -N <rather start with a low number to test>
#SBATCH -n <N * cores / node>
 
module purge
 
module load chem/NAMD
 
# the suffix 'namd_conf' is arbitrary
srun namd2  <prefix>.namd_conf

The configuration file content depends on the simulation to be carried out, however, the coordinates and structure parameters could simply refer to the auto-generated .pdf and .psf files produced by VMD as described, respectively.

The NAMD tutorial indicates this download site for CHARMM force fields and topology files³⁾

When using these force field files, be sure not to forget to amend the force field type to the configuration file, e.g.:

parameters           <path to desired prm>.prm
paraTypeCharmm       on