[gmx-users] dt in mdp

[Justin Lemkul]

On 4/10/20 8:13 AM, Alex wrote:
> Dear Justin,
> Any comment please?

Sorry, haven't had power/network for a while due to some bad storms here.

GROMOS force fields are parametrized assuming all bonds are fixed, so 
your constraints should be "all-bonds" not "h-bonds."

I would also suggest you thoroughly validate the quality of the epoxy 
molecule topology against QM data and bulk-phase properties, if possible.

-Justin

> Regards,
> Alex
>
> On Tue, Apr 7, 2020 at 5:38 PM Alex <alexanderwien2k at gmail.com> wrote:
>
>> Thanks Justin for the response.
>> Please find below the mdp file.
>> The system is a thin film made out of a epoxy molecule, picture in be
>> below link, with water on top and bottom of the film. I even sometimes have
>> the same issue when I simulate the bulk of this system, I mean a cubic box
>> filled by this molecule and no water.
>> I got the force fields from the latest version of ATB repository by which
>> I have previously done some other simulation for the same molecule.
>>
>> https://drive.google.com/open?id=1tJLxh9jQ2v5DDTrVR8IuQz40B3NN0zlb
>>
>> %--------mdp-------
>> title                    = Thin-Film
>> integrator               = md
>> dt                       = 0.002     ; 2 fs      ;0.001 1 fs
>> nsteps                   = 25000000   ; 50 ns      ;50000000   ; 50 ns
>> xtc-precision            = 500       ; 1000
>> ;nstlist                  = 40
>> %-----------------
>> ;;in .trr file
>> nstxout                  = 3000      ; 6000
>> nstvout                  = 0
>> nstfout                  = 0
>> ;;in energy file.log
>> nstlog                   = 2000       ; 4000
>> nstcalcenergy            = 2000       ;4000
>> nstenergy                = 2000         ;4000
>> ;;in xtc file
>> nstxout-compressed       = 0
>> ;compressed-x-grps        = non-Water
>> %-----------------
>> continuation             = yes
>> gen-vel                  = no
>> constraint-algorithm     = lincs
>> constraints              = h-bonds
>> cutoff-scheme            = Verlet
>> coulombtype              = PME
>> rcoulomb                 = 1.4
>>
>> vdwtype                  = Cut-off
>> rvdw                     = 1.4
>> DispCorr                 = EnerPres
>>
>> tcoupl                    = v-rescale
>> tc-grps                   = system
>> tau-t                     = 1.5
>> nhchainlength             = 10
>> ref-t                     = 298.15
>> pbc                       = xyz
>>
>> pcoupl                   = Parrinello-Rahman
>> Pcoupltype               = isotropic
>> tau_p                    = 2.5
>> compressibility          = 4.5e-5
>> ref_p                    = 1.0
>> refcoord-scaling         = com
>> energygrps               = thin_film SOL
>> comm-mode                = Linear
>> nstcomm                  = 100
>> comm-grps                = Thin_fiml SOL
>> %-------------------------
>>
>> Thank you
>> Alex
>>
>> On Tue, Apr 7, 2020 at 5:06 PM Justin Lemkul <jalemkul at vt.edu> wrote:
>>
>>>
>>> On 4/7/20 5:00 PM, Alex wrote:
>>>> Dear all,
>>>> After minimization and equalizations using nvt (v-rescale) and npt (both
>>>> berendsen and ;Parrinello-Rahman), a simulation of mine could run well
>>> for
>>>> 200 ns using dt =0.001 while it would crash after 3 ns If I used dt =
>>> 0.002
>>>> with the particles communication fatal error.
>>>>
>>>> Fatal error:
>>>> 2 particles communicated to PME rank 12 are more than 2/3 times the
>>> cut-off
>>>> out of the domain decomposition cell of their charge group in dimension
>>> y.
>>>> This usually means that your system is not well equilibrated.
>>>>
>>>> So, if the system would not be well equlibrated, then I would expect
>>> that
>>>> with dt = 0.001 the simulation wouldn't run well for 200 ns. But I
>>> expect
>>>> that it also crashes for example around 6 ns as the with the dt = 0.002
>>> the
>>>> simulation last only 3 ns.
>>>>
>>>> Any comment that helps to understand the problem would be highly
>>>> appreciated.
>>> Please provide a description of what your system is and a full .mdp
>>> file. While most of the time these crashes come from poor equilibration,
>>> an inadequately parametrized topology or bad combination/misuse of
>>> algorithms can also cause crashes.
>>>
>>> -Justin
>>>
>>> --
>>> ==================================================
>>>
>>> Justin A. Lemkul, Ph.D.
>>> Assistant Professor
>>> Office: 301 Fralin Hall
>>> Lab: 303 Engel Hall
>>>
>>> Virginia Tech Department of Biochemistry
>>> 340 West Campus Dr.
>>> Blacksburg, VA 24061
>>>
>>> jalemkul at vt.edu | (540) 231-3129
>>> http://www.thelemkullab.com
>>>
>>> ==================================================
>>>
>>> --
>>> 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.
>>>

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

Justin A. Lemkul, Ph.D.
Assistant Professor
Office: 301 Fralin Hall
Lab: 303 Engel Hall

Virginia Tech Department of Biochemistry
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.thelemkullab.com

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