[gmx-users] NPT, freezegroups and Position Restraints.
Mark Abraham
mark.j.abraham at gmail.com
Thu Aug 22 12:20:19 CEST 2013
On Thu, Aug 22, 2013 at 12:51 AM, HANNIBAL LECTER
<hanniballecter13 at gmail.com> wrote:
> Thanks. I am not interested in the real physics of the freezegroups as I am
> assuming the fluctuations are negligible in the timescale of the
> fluctuations of the non-freeze groups.
That may be true in the case where the non-interesting group is
unfrozen, but it does not follow that the perturbation to the
Hamiltonian from freezing the atoms (ie. that all interactions with
them now don't follow Newton's third law) is equally negligible.
> I thought that the langevin
> thermostat is quite effective in this case just to ensure that there is no
> net heat flow from the hot to the cold freeze groups. I maybe missing
> something.
There's no heat flow at all to the frozen groups because they have no
velocity, so no temperature and no coupling. The atoms near the frozen
groups are going to be weird, and that has flow-on effects. Whether
those effects are significant is up to you to show, but you should
have a much easier time setting up a stable non-frozen CNT. Freezing a
group does not make your simulation appreciably faster. If I were to
review a paper that used freeze groups and made no attempt to show the
results were unaffected, then that paper would be on the fast track
for rejection... Relying on the thermostat to mask errors in the model
physics or the integration of it is similarly reprehensible, IMO. :-)
Mark
> On Wed, Aug 21, 2013 at 6:03 PM, Mark Abraham <mark.j.abraham at gmail.com>wrote:
>
>> On Wed, Aug 21, 2013 at 10:22 PM, HANNIBAL LECTER
>> <hanniballecter13 at gmail.com> wrote:
>> > I am not sure what is the problem with using NPT, constraints and
>> > freezegroups. Unless there is some fundamental issue (which I cannot
>> think
>> > of) one should in principle be able to run NPT with the CNT system by
>> > freezing the CNT group.
>>
>> Frozen atoms do not have velocity, and the energy they would have
>> acquired in the update step is simply removed. So you get a
>> discontinuity every time step, because the new configuration did not
>> evolve from the previous one in a way that could mimic the statistics
>> of the ensemble with no frozen atoms. Also, there must be net flow of
>> heat if the temperature is constant in (even just) NVT, so you're not
>> modelling a normal ensemble at all. Constraints and box-size rescaling
>> in NPT make things worse. I can't imagine a use case (other than
>> equilibration of a problematic system) where you could use freeze
>> groups and believe you were modelling something like real physics.
>>
>> > I am using sd with LINCS to constraint the bonds.
>> > However, if md is used with SHAKE the system crashes with segmentation
>> > fault. Does anyone have any idea as to what could be the problem?
>>
>> Using freeze groups.
>>
>> Mark
>>
>> > On Mon, Aug 19, 2013 at 6:40 PM, HANNIBAL LECTER <
>> hanniballecter13 at gmail.com
>> >> wrote:
>> >
>> >> Interestingly, removing position restraints does not have much of an
>> >> impact.
>> >>
>> >>
>> >> On Mon, Aug 19, 2013 at 1:40 PM, Justin Lemkul <jalemkul at vt.edu> wrote:
>> >>
>> >>>
>> >>>
>> >>> On 8/19/13 11:58 AM, HANNIBAL LECTER wrote:
>> >>>
>> >>>> I usually use 1.0 for my systems and I get the correct values for the
>> >>>> temperatures. If I couple it as a system, the low T-CNT yields a
>> >>>> different
>> >>>> value than the target 300K.
>> >>>>
>> >>>>
>> >>> What about changing the integrator to md? What happens if you remove
>> >>> restraints while using sd?
>> >>>
>> >>> -Justin
>> >>>
>> >>>
>> >>>> On Mon, Aug 19, 2013 at 11:48 AM, Justin Lemkul <jalemkul at vt.edu>
>> wrote:
>> >>>>
>> >>>>
>> >>>>>
>> >>>>> On 8/19/13 11:42 AM, HANNIBAL LECTER wrote:
>> >>>>>
>> >>>>> Thanks Justin.
>> >>>>>>
>> >>>>>> I am not sure how it is fundamentally incompatible. Especially with
>> >>>>>> refcoord-scaling = com. Can you please provide with some more input
>> on
>> >>>>>> that?
>> >>>>>>
>> >>>>>>
>> >>>>>> The refcoord-scaling option doesn't matter with frozen groups.
>> >>>>> Anything
>> >>>>> that is frozen, by definition, never has its position updated. Under
>> >>>>> the
>> >>>>> influence of pressure coupling, other particles around the frozen
>> group
>> >>>>> can
>> >>>>> have their positions scaled and thus collide with the frozen group,
>> >>>>> which
>> >>>>> has remained in its original location.
>> >>>>>
>> >>>>>
>> >>>>>
>> >>>>> The outputs are from g_energy.
>> >>>>>>
>> >>>>>>
>> >>>>>> OK, good.
>> >>>>>
>> >>>>>
>> >>>>> Here is the .mdp file.
>> >>>>>
>> >>>>>>
>> >>>>>> define = -DPOSRES_CNT
>> >>>>>> constraints = all-bonds
>> >>>>>> pbc = xyz
>> >>>>>> integrator = sd
>> >>>>>> ld_seed = 215546
>> >>>>>> dt = 0.002 ; ps !
>> >>>>>> nsteps = 50000 ; 50 ns
>> >>>>>> ;nsteps = 125000 ; 250 ps
>> >>>>>> nstcomm = 10
>> >>>>>> nstcalcenergy = 10
>> >>>>>> nstxout = 0
>> >>>>>> nstxtcout = 5000 ;every 10 ps
>> >>>>>> nstvout = 100
>> >>>>>> nstfout = 0
>> >>>>>> nstlog = 1000
>> >>>>>> nstenergy = 1000
>> >>>>>> nstlist = 10
>> >>>>>> ns_type = grid
>> >>>>>> rlist = 1.0
>> >>>>>> vdwtype = cut-off
>> >>>>>> rvdw = 1.0
>> >>>>>> coulombtype = pme
>> >>>>>> rcoulomb = 1.0
>> >>>>>> fourierspacing = 0.12
>> >>>>>> pme_order = 4
>> >>>>>> ewald_rtol = 1e-5
>> >>>>>> optimize_fft = yes
>> >>>>>> tc-grps = Protein CNT SOL
>> >>>>>> tau_t = 1.0 1.0 1.0
>> >>>>>>
>> >>>>>>
>> >>>>> Note that tau_t is the inverse friction constant, not the "normal"
>> >>>>> coupling constant. In this case, I don't know what appropriate
>> values
>> >>>>> are,
>> >>>>> especially if one of your groups is restrained. The manual advises a
>> >>>>> value
>> >>>>> of 2.0, so at least try that. Otherwise, try coupling the System as
>> a
>> >>>>> whole or use a different integrator (i.e., md) to try to figure out
>> the
>> >>>>> origin of the problem.
>> >>>>>
>> >>>>> -Justin
>> >>>>>
>> >>>>> ref_t = 300.000 300 300
>> >>>>>
>> >>>>>> ;freezegrps = CNT
>> >>>>>> ;freezedim = Y Y Y
>> >>>>>> ; Energy monitoring
>> >>>>>> energygrps = CNT Protein SOL
>> >>>>>> ; energygrp-excl = CNT CNT
>> >>>>>> ; Isotropic pressure coupling is now on
>> >>>>>> Pcoupl = Berendsen
>> >>>>>> refcoord-scaling = com
>> >>>>>> Pcoupltype = isotropic
>> >>>>>> tau_p = 5
>> >>>>>> compressibility = 4.5e-5
>> >>>>>> ref_p = 1.0
>> >>>>>> gen_vel = yes
>> >>>>>> gen_temp = 300.000
>> >>>>>> gen_seed = 981487
>> >>>>>>
>> >>>>>>
>> >>>>>>
>> >>>>>> On Mon, Aug 19, 2013 at 11:35 AM, Justin Lemkul <jalemkul at vt.edu>
>> >>>>>> wrote:
>> >>>>>>
>> >>>>>>
>> >>>>>>
>> >>>>>>> On 8/19/13 11:32 AM, HANNIBAL LECTER wrote:
>> >>>>>>>
>> >>>>>>> Hi,
>> >>>>>>>
>> >>>>>>>>
>> >>>>>>>> I have been performing NPT simulations with CNT, Protein and
>> water.
>> >>>>>>>> However, when I freeze the CNT, the system crashes with
>> segmentation
>> >>>>>>>> fault.
>> >>>>>>>>
>> >>>>>>>>
>> >>>>>>>> That is to be expected. Frozen atoms and pressure coupling are
>> >>>>>>>>
>> >>>>>>> fundamentally incompatible.
>> >>>>>>>
>> >>>>>>>
>> >>>>>>> If I use position restraints, the temperature of the system is
>> >>>>>>> lower
>> >>>>>>> than
>> >>>>>>>
>> >>>>>>> what it is expected to be. I am using sd coupling.
>> >>>>>>>>
>> >>>>>>>> If I couple the protein, CNT and SOL separately, the temperature
>> of
>> >>>>>>>> the
>> >>>>>>>> peptide is lower than 300K (~395K) while the temperature of the
>> CNT
>> >>>>>>>> is
>> >>>>>>>> 157K. The target temperature is 300K in all cases. Can any please
>> >>>>>>>> explain
>> >>>>>>>> why such discrepancies are occurring?
>> >>>>>>>>
>> >>>>>>>>
>> >>>>>>>> Are these temperatures from g_energy or g_traj? Can you please
>> >>>>>>>> post a
>> >>>>>>>>
>> >>>>>>> full .mdp file?
>> >>>>>>>
>> >>>>>>> -Justin
>> >>>>>>>
>> >>>>>>> --
>> >>>>>>> ==============================******====================
>> >>>>>>>
>> >>>>>>>
>> >>>>>>>
>> >>>>>>> Justin A. Lemkul, Ph.D.
>> >>>>>>> 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 <jalemkul at outerbanks.**
>> >>>>>>> umaryland.edu <jalemkul at outerbanks.**umaryland.edu<
>> jalemkul at outerbanks.umaryland.edu>>>
>> >>>>>>> | (410)
>> >>>>>>> 706-7441
>> >>>>>>>
>> >>>>>>> ==============================******====================
>> >>>>>>>
>> >>>>>>>
>> >>>>>>> --
>> >>>>>>> gmx-users mailing list gmx-users at gromacs.org
>> >>>>>>> http://lists.gromacs.org/******mailman/listinfo/gmx-users<
>> http://lists.gromacs.org/****mailman/listinfo/gmx-users>
>> >>>>>>> <htt**p://lists.gromacs.org/****mailman/listinfo/gmx-users<
>> http://lists.gromacs.org/**mailman/listinfo/gmx-users>
>> >>>>>>> >
>> >>>>>>> <htt**p://lists.gromacs.org/**mailman/**listinfo/gmx-users<
>> http://lists.gromacs.org/mailman/**listinfo/gmx-users>
>> >>>>>>> <h**ttp://lists.gromacs.org/**mailman/listinfo/gmx-users<
>> http://lists.gromacs.org/mailman/listinfo/gmx-users>
>> >>>>>>> >
>> >>>>>>>
>> >>>>>>>
>> >>>>>>>> * Please search the archive at http://www.gromacs.org/**
>> >>>>>>> Support/Mailing_Lists/Search<**h**ttp://
>> www.gromacs.org/**Support/**<http://www.gromacs.org/Support/**>
>> >>>>>>>
>> >>>>>>> Mailing_Lists/Search<http://**www.gromacs.org/Support/**
>> >>>>>>> Mailing_Lists/Search<
>> http://www.gromacs.org/Support/Mailing_Lists/Search>
>> >>>>>>> >>before
>> >>>>>>> posting!
>> >>>>>>>
>> >>>>>>> * Please don't post (un)subscribe requests to the list. Use the www
>> >>>>>>> interface or send it to gmx-users-request at gromacs.org.
>> >>>>>>> * Can't post? Read
>> http://www.gromacs.org/******Support/Mailing_Lists<
>> http://www.gromacs.org/****Support/Mailing_Lists>
>> >>>>>>> <http://**www.gromacs.org/**Support/**Mailing_Lists<
>> http://www.gromacs.org/**Support/Mailing_Lists>
>> >>>>>>> >
>> >>>>>>> <http://**www.gromacs.org/**Support/**Mailing_Lists<
>> http://www.gromacs.org/Support/**Mailing_Lists>
>> >>>>>>> <http:/**/www.gromacs.org/Support/**Mailing_Lists<
>> http://www.gromacs.org/Support/Mailing_Lists>
>> >>>>>>> >
>> >>>>>>>
>> >>>>>>>>
>> >>>>>>>>
>> >>>>>>>
>> >>>>>>> --
>> >>>>> ==============================****====================
>> >>>>>
>> >>>>> Justin A. Lemkul, Ph.D.
>> >>>>> 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 <jalemkul at outerbanks.**
>> >>>>> umaryland.edu <jalemkul at outerbanks.umaryland.edu>> | (410)
>> >>>>> 706-7441
>> >>>>>
>> >>>>> ==============================****====================
>> >>>>> --
>> >>>>> gmx-users mailing list gmx-users at gromacs.org
>> >>>>> http://lists.gromacs.org/****mailman/listinfo/gmx-users<
>> http://lists.gromacs.org/**mailman/listinfo/gmx-users>
>> >>>>> <htt**p://lists.gromacs.org/mailman/**listinfo/gmx-users<
>> http://lists.gromacs.org/mailman/listinfo/gmx-users>
>> >>>>> >
>> >>>>> * Please search the archive at http://www.gromacs.org/**
>> >>>>> Support/Mailing_Lists/Search<h**ttp://www.gromacs.org/Support/**
>> >>>>> Mailing_Lists/Search<
>> http://www.gromacs.org/Support/Mailing_Lists/Search>>before
>> >>>>> posting!
>> >>>>> * Please don't post (un)subscribe requests to the list. Use the www
>> >>>>> interface or send it to gmx-users-request at gromacs.org.
>> >>>>> * Can't post? Read http://www.gromacs.org/****Support/Mailing_Lists<
>> http://www.gromacs.org/**Support/Mailing_Lists>
>> >>>>> <http://**www.gromacs.org/Support/**Mailing_Lists<
>> http://www.gromacs.org/Support/Mailing_Lists>
>> >>>>> >
>> >>>>>
>> >>>>>
>> >>> --
>> >>> ==============================**====================
>> >>>
>> >>> Justin A. Lemkul, Ph.D.
>> >>> 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 <jalemkul at outerbanks.umaryland.edu>|
>> (410)
>> >>> 706-7441
>> >>>
>> >>> ==============================**====================
>> >>> --
>> >>> gmx-users mailing list gmx-users at gromacs.org
>> >>> http://lists.gromacs.org/**mailman/listinfo/gmx-users<
>> http://lists.gromacs.org/mailman/listinfo/gmx-users>
>> >>> * Please search the archive at http://www.gromacs.org/**
>> >>> Support/Mailing_Lists/Search<
>> http://www.gromacs.org/Support/Mailing_Lists/Search>before posting!
>> >>> * Please don't post (un)subscribe requests to the list. Use the www
>> >>> interface or send it to gmx-users-request at gromacs.org.
>> >>> * Can't post? Read http://www.gromacs.org/**Support/Mailing_Lists<
>> http://www.gromacs.org/Support/Mailing_Lists>
>> >>>
>> >>
>> >>
>> > --
>> > gmx-users mailing list gmx-users at gromacs.org
>> > http://lists.gromacs.org/mailman/listinfo/gmx-users
>> > * Please search the archive at
>> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
>> > * Please don't post (un)subscribe requests to the list. Use the
>> > www interface or send it to gmx-users-request at gromacs.org.
>> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>> --
>> gmx-users mailing list gmx-users at gromacs.org
>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>> * Please search the archive at
>> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
>> * Please don't post (un)subscribe requests to the list. Use the
>> www interface or send it to gmx-users-request at gromacs.org.
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
> --
> gmx-users mailing list gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> * Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
More information about the gromacs.org_gmx-users
mailing list