[gmx-users] A question about "blowing up"
jalemkul at vt.edu
Sat Mar 1 12:59:30 CET 2014
On 3/1/14, 1:14 AM, sujithkakkat . wrote:
> I am following justin lemkul's tutorials for the cyclohexane-water
> biphasic system using Gromos96 forcefield.
> Things went well till NVT equilibration where I had this "Blowing up" issue.
> Shortening the time step alone did not help, and it was difficult for me
> to figure out what was happening from watching the trajectory.
> g_energy data indicated a spike on the, short range LJ contribution at
> around 10ps.
> Then I switched off the constraints (but not for water, SETTLE still
> used) and ran the simulation again with a small time step. Things worked
> well this time , the system reaching the target temperature.
> I want to make sure that switching off the constraints is a reasonable
> thing to solve Blowing up and that it would not affect in any way the
> accuracy of further simulations.
> http://www.gromacs.org/Documentation/Terminology/Blowing_Up says that
> constraints are exactly not the source of the Blowing up problem.
> suggests removing constraints as a solution.
> So, in general, is the problem solved by removing constrains, or am I
> cheating? Please comment.
Constraints are not a problem, they are a symptom. It is not obligatory to use
constraints, so the advice on the page above is somewhat generic in that if you
are not using constraints you should also be decreasing the timestep (usually
obligatory), but then you've changed two variables and it is not clear which one
was the solution.
> Here is the .mdp file which solved the Blowing up issue.
> ; 7.3.3 Run Control
> integrator = md
> tinit = 0
> dt = 0.0002
> nsteps = 2500000
> comm_mode = Linear
> nstcomm = 1
> comm_grps = CHX SOL
> ; 7.3.8 Output Control
> nstxout = 25000
> nstvout = 25000
> nstfout = 25000
> nstlog = 100
> nstenergy = 100
> nstxtcout = 100
> xtc_precision = 1000
> xtc_grps = System
> energygrps = System
> ; 7.3.9 Neighbor Searching
> nstlist = 1
You lose serious performance by doing this. Setting nstlist = 10 is sufficient
for MD. Using nstlist = 1 is only for energy minimization.
> ns_type = grid
> pbc = xyz
> rlist = 0.9
> ; 7.3.10 Electrostatics
> coulombtype = PME
> rcoulomb = 0.9
> ; 7.3.11 VdW
> vdwtype = cut-off
> rvdw = 1.4
> DispCorr = EnerPres
> ; 7.3.13 Ewald
> fourierspacing = 0.12
> pme_order = 4
> ewald_rtol = 1e-5
> ; 7.3.14 Temperature Coupling
> tcoupl = berendsen
> tc_grps = CHX SOL
> tau_t = 0.1 0.1
> ref_t = 298 298
> ; 7.3.17 Velocity Generation
> gen_vel = yes
> gen_temp = 298
> gen_seed = -1
> ; 7.3.18 Bonds
> constraints = none
> constraint_algorithm = LINCS
> continuation = no
> lincs_order = 4
> lincs_iter = 1
> lincs_warnangle = 30
Nothing in the .mdp file stands out as being the source of the problem. How
successful was energy minimization? What were Epot and Fmax?
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