[gmx-users] Still having issues analyzing normal-mode hessian

[Justin Lemkul]

On 5/6/13 2:16 PM, Bryan Roessler wrote:
> Hello,
>
> I have successfully generated a 945MB mtx hessian storing the normal modes
> of a ~1500AA protein using mdrun with the "nm" integrator and no cutoffs.
> However, when I try to analyze the normal modes and create trajectories
> using g_nmeig, I can't seem to generate any output. I am running
> g_nmeig_mpi_d  with the following options: "mpiexec_mpt g_nmeig_mpi_d -f
> input.mtx -s NMA.tpr -of NMA_freq.xvg -ol NMA_val.xvg -v NMA_vec.trr" and

There is no point in trying to run in parallel; only mdrun is parallelized. 
That could be one source of problem.

> have been letting it run a few times for up to 5 days on 128 Itanium nodes
> with 120gb maximum memory. When the job finally aborts after running out of
> cpu time, I get a standard output file that's about 8GB (which I am unable
> to open). Is there a way to force g_nmeig to generate some output or at
> least show me what is going on? Am I running it with the correct options?
> There isn't a lot of documentation on these normal-mode specific programs
> so I don't know how to begin troubleshooting.
>

Run interactively and try with a smaller data set, something that will consume 
very little memory, then scale up.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================