[gmx-users] Error:EM did not converge while using polarizable ion
折晓会
shexh19 at hotmail.com
Tue Aug 4 00:38:08 CEST 2015
Dear professor Justin,
Thank you very much for your suggestions! I really appreciate your help!
I build a larger system and use the parameters you suggested. The simulation is working better now, especially after decreasing the time step, but EM occasionally does not converge in some steps during the simulation. For example,
step 341262: EM did not converge in 100 iterations, RMS force 2.608
step 341263: EM did not converge in 100 iterations, RMS force 15.730
step 341264: EM did not converge in 100 iterations, RMS force 40.663
step 341265: EM did not converge in 100 iterations, RMS force 88.519
step 341266: EM did not converge in 100 iterations, RMS force 173.432
step 473594: EM did not converge in 100 iterations, RMS force 9.094
step 473595: EM did not converge in 100 iterations, RMS force 24.042
step 473596: EM did not converge in 100 iterations, RMS force 57.226
step 473597: EM did not converge in 100 iterations, RMS force 127.469
step 473598: EM did not converge in 100 iterations, RMS force 285.591
step 473599: EM did not converge in 100 iterations, RMS force 479.862
I am wondering if this is common for polarizable model simulations. I tried to use different time steps (0.2, 0.5 and 0.8 fs) and longer steps (500 000, 800 000), but the similar warnings occured periodically as above.
In addition, I am also confused about the outcome of the energy minimization, since it reaches neither the emtol (1) nor the nsteps (50000):
Energy minimization has stopped, but the forces have not converged to the
requested precision Fmax < 1 (which may not be possible for your system). It
stopped because the algorithm tried to make a new step whose size was too
small, or there was no change in the energy since last step.
Steepest Descents converged to machine precision in 2115 steps,
but did not reach the requested Fmax < 1.
Potential Energy = -1.0560062e+09
Maximum force = 1.7716933e+15 on atom 2483
Norm of force = 4.9991217e+13
Based on this outcome, is it due to the low machine precision which stops the energy minimization?
Best wishes,
Xiaohui
> Date: Sat, 1 Aug 2015 19:19:58 -0400
> From: jalemkul at vt.edu
> To: gmx-users at gromacs.org
> Subject: Re: [gmx-users] Error:EM did not converge while using polarizable ion
>
>
>
> On 8/1/15 6:38 PM, 折晓会 wrote:
> > Thank you Professor Justin for your reply!
> >
> > The .mdp files for energy minimization and equilibration are shown in the
> > following. During the energy minimization, the simulation stops at step=38,
> > which may be unreasonable. The initial configuration of the system is simple
>
> Well, what is the maximum force after EM? That tells you whether or not its
> reasonable, not the number of steps taken.
>
> > cubic. The drude particle and its core atom are placed at the same position.
>
> Yep, I know how Drude systems are built; I develop the force field :)
>
> >
> > The .mdp file for energy minimization is:
> >
> > integrator = steep emtol = 1 niter = 100 emstep =
> > 0.01 nsteps = 50000
> >
> > nstcalcenergy = 1 constraints = none
> >
> > nstlist = 1 cutoff-scheme = Verlet ns_type =
> > grid coulombtype = PME rcoulomb = 0.8 rvdw =
> > 0.8 pbc = xyz
> >
>
> These cutoffs are much too short.
>
> > The outcome of the energy minimization is:
> >
> > Step= 31, Dmax= 2.7e-05 nm, Epot= -2.06027e+07 Fmax= 3.15846e+12, atom=
> > 600 Step= 35, Dmax= 4.1e-06 nm, Epot= -1.74228e+08 Fmax= 2.28102e+14, atom=
> > 600 Step= 38, Dmax= 1.2e-06 nm, Epot= -9.57138e+07 Fmax= 6.87661e+13, atom=
> > 600 Energy minimization has stopped, but the forces have not converged to
> > the requested precision Fmax < 1 (which may not be possible for your system).
> > It stopped because the algorithm tried to make a new step whose size was too
> > small, or there was no change in the energy since last step. Either way, we
> > regard the minimization as converged to within the available machine
> > precision, given your starting configuration and EM parameters.
> >
> > The .mdp file for equilibration is:
> >
>
> Lots of problems here. See annotations below.
>
> > dt = 0.002
> >
>
> dt should never be larger than 0.001 for a polarizable system, *especially* one
> with ions. Water is somewhat forgiving; ions and biomolecules are not. This is
> probably your biggest problem, but you have lots more.
>
> > ; VDW vdwtype = cut-off
> > vdw-modifier = Potential-shift
> > rvdw = 0.8
> > DispCorr = EnerPres
> >
>
> Use:
>
> vdwtype = cutoff
> vdw-modifier = force-switch
> rvdw-switch = 1.0
> rvdw = 1.2
> DispCorr = EnerPres
>
> > ; Electrostatics coulombtype = PME
> > coulomb-modifier = Potential-shift-Verlet
> > rcoulomb = 0.8
>
> Use rcoulomb = 1.2
>
> > pme_order = 4
> > fourierspacing = 0.8
>
> Wildly inappropriate fourierspacing. Your PME will be terribly inaccurate. Use
> 0.1 or 0.12.
>
> >
> > ; Temperature coupling is on
> > tcoupl = nose-hoover
>
> It is unwise to start the simulation with Nose-Hoover in the SCF regime. Try
> something more forgiving like V-rescale.
>
> > nh-chain-length = 1
> > tc-grps = Water non-Water
>
> Too few degrees of freedom in the ion group to justify its own group. Just
> couple System.
>
> > tau_t = 0.8 0.8
>
> You seem to like 8's :) This may be too weak of coupling. Try 0.5 or 0.1.
>
> -Justin
>
> --
> ==================================================
>
> 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
> http://mackerell.umaryland.edu/~jalemkul
>
> ==================================================
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.
More information about the gromacs.org_gmx-users
mailing list