[gmx-users] Truncated .trr file size - 20% smaller than expected
jalemkul at vt.edu
Fri May 9 19:53:16 CEST 2014
On 5/9/14, 1:46 PM, Douglas Grahame wrote:
> Hi there this is my first time posting here so my apologies if I disregard
> any general formatting rules! In any event on to my question/problem.
> I am fairly new to MD simulations with much more background on the
> biological side than the simulation side. That being said, I have been
> receiving training from another member in the lab and have just started my
> simulations of a 325 residue protein using GROMACS 4.5.5. My issue has come
> from the fact that when I grompp my EM.mdp, NVT.mdp, or MPT.mdp files
> everything seems to be fine ie. No errors or issues and the file sizes match
> the expected output size. However, when I run my simulation I'm getting
> about a %20 decrease in the size of my .trr file from what it expects.
> From my .out file for the molecular dynamics run I get 'WARNING: This run
> will generate roughly 24940 Mb of data'
> Actual amount of data produced is ~18 Gb of data.
> There are no errors running gmx_check on any of the tpr files are anything
> I've tried running the same simulation on two different supercomputing
> clusters to see if the same error occurs just in case it was an issue of the
> build version having any effect on the size and run into the same problems
> I've tried running the simulation in parts just in case there was some
> memory issue causing some problem that I wasn't aware of or looking for
> I've gotten my labmate and one more person familiar with GROMACS to look at
> my issued commands and they seem stumped
> I've also tried running it on 4.6.3 with the verlet cutoff scheme as I was
> having some issues with pp/pme load balance
> Nothing so far has made a difference and although the data seems to be fine
> I'm loathe to do all the analysis if there is something majorly wrong with
> my simulation. Problem for me is I don't know if or by how much the actual
> trr file size should change from the expected amounts.
> Command submitted: mdrun_mpi -npme -1 -s MD.tpr -o MD.trr -x MD.xtc -c
> MD.pdb -e MD.edr -g MD.log -v -dlb yes -cpo MD.cpt
For what it's worth, save yourself some typing with:
mdrun_mpi -npme -1 -deffnm MD -dlb yes
> I am running/wanting to run my simulations for 25 - 50ns
Probably the grompp estimate is just off. If gmxcheck confirms you have
everything you should, there's no reason to think there is a problem.
Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow
Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201
jalemkul at outerbanks.umaryland.edu | (410) 706-7441
More information about the gromacs.org_gmx-users