[gmx-users] Various errors in pulling simulations
jalemkul at vt.edu
Fri Feb 20 14:57:58 CET 2015
On 2/20/15 8:25 AM, Tim Stauch wrote:
> Dear all,
> I want to use GROMACS to unfold a protein (1TIT.pdb). After the energy minimization and the equilibration runs, I want to run my production simulations, but I get different fatal errors, depending on my setup, e.g.:
> "Fatal error:
> 9 particles communicated to PME node 1 are more than 2/3 times the cut-off out of the domain decomposition cell of their charge group in dimension x.
> This usually means that your system is not well equilibrated."
> "Fatal error:
> A charge group moved too far between two domain decomposition steps
> This usually means that your system is not well equilibrated"
> One of these errors occurs suddenly in the middle of the simulation after several hundred steps. However, I have found out that the errors occur the earlier the higher the applied force is. This is why I don’t think that it is a problem with the equilibration.
> I’ve decreased the time step (dt) to half a femtosecond, but even then the errors occur.
> The mdp options for the pull command look like this:
> pull = umbrella
> pull_geometry = distance
> pull_dim = Y N N
> pull_start = yes
> pull_ngroups = 1
> pull_group0 = punkt1
> pull_group1 = punkt2
> pull_rate1 = 0.01
> pull_k1 = (I’m varying this between 1000 and 5000)
> I have defined “punkt1” and “punkt2” in an index file (atoms 1 and 1377, i.e. the terminal N and C atoms of the molecule).
> I am using GROMACS 4.5.5.
> I would really appreciate if you could help me eliminate these errors, so that I can unfold my protein.
How about a slower pull rate? It sounds like things are just colliding in a
very nasty way and things are blowing up.
Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow
Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
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