[gmx-users] Reg: NPT Equilibration of water
Mark Abraham
Mark.Abraham at anu.edu.au
Sun Oct 31 12:24:46 CET 2010
On 31/10/2010 8:17 PM, Tsjerk Wassenaar wrote:
> Hey,
>
> There's quite a number of people that should be rereading their statistics... :p
>
> First of all, there are instantaneous measurements, averages and
> fluctuations. If the statistics (mean/fluctuation) are 7 +/- 500, then
> that doesn't mean that the average of 7 is an estimate that may be off
> by 500, and the true average can lie between -493 and 507. It means
> that the average is 7, but that any instantaneous measurement may be
> expected to lie between those values, and that two measurements, say
> -300 and +300, can not be expected to be from different distributions.
> So a pressure of 300 bar measured in one frame does not suggest that
> the pressure of the system is higher than that of a system of the same
> constitution with a pressure measured in one frame of -300 bar.
Indeed, I was speaking crudely :-) One wants to measure the actual
standard error (or similar) of average pressure observations over a
given length of time to have proper confidence bounds for the estimate
of the average pressure.
> Now the real issue is that the average value (7) is an estimate of the
> true mean, which is expected to be 1, as that is given as target
> value. The average was determined over a short time, 100ps. Every
> interval of 100ps will yield a different value for the observation
> 'average over 100ps', and these averages will have a certain
> distribution around an average 'average over 100ps' observations. If
> you do ten of these then the average over averages will still only be
> an average over 1 ns, which will be a better estimator of the true
> mean value, but is still an estimate itself. And this can be continued
> ad infinitum. The average observed over a certain time interval will
> converge to the true mean with increasing interval length.
>
> The standard deviation of 'average pressures observed over 100ps' is
> likely more than a few bar if the fluctuation in instantaneous values
> is so huge. In that regard, an observed average pressure of 7 bar over
> 100ps is probably consistent with a system at a mean pressure of 1
> bar.
Yes.
Mark
> Hope it helps,
>
> Tsjerk
>
> On Fri, Oct 29, 2010 at 4:58 PM, Mark Abraham<Mark.Abraham at anu.edu.au> wrote:
>> On 29/10/2010 7:50 PM, vinothkumar mohanakrishnan wrote:
>>
>> Hi Mark
>>
>> How come 7 +/- 500 bar is approximately 1 bar. Can you explain it more
>> clearly? It will be of more useful to me to understand the concept.
>>
>> It's a pretty basic one in scientific measurement. See
>> http://en.wikipedia.org/wiki/Error_bar and links thereon, or an introductory
>> physics text.
>>
>> When we make a measurement, that's only an estimate of the true value. When
>> someone cites a range of uncertainty, they are acknowledging that there is
>> an appreciable chance that the true value, of which they have taken an
>> error-prone measurement, lies somewhere in a given range. So if 7 +/-500
>> acknowledges that the true value could lie anywhere from about -493 to 507,
>> then that's not much less useful than an observation of 1 +/- 500, which
>> acknowledges that the true value could lie anywhere from about -499 to 501.
>> The details of that chance will vary with exactly what variation quantity is
>> being reported (standard deviation, standard error, maximum deviation,
>> quartile, etc.).
>>
>> Here, g_energy is effectively reporting that the instantaneous pressure
>> values varied over a range of many hundreds of bar, and that you haven't
>> made enough observations to report a reliable average.
>>
>> Mark
>>
>>
>> One more thing, density (from g_energy command) of water is found to low
>> (expected 1000) after NPT is equilibration. why? given below is my average
>> density value. Can i proceed with this density value?
>>
>> Energy Average RMSD Fluct. Drift
>> Tot-Drift
>> -------------------------------------------------------------------------------
>> Density (SI) 979.37 14.3238 14.2431 -0.0525784
>> -5.25789
>>
>> Regards
>> Vinoth
>>
>> On Fri, Oct 29, 2010 at 2:06 PM, Mark Abraham<Mark.Abraham at anu.edu.au>
>> wrote:
>>> On 29/10/2010 6:39 PM, vinothkumar mohanakrishnan wrote:
>>>
>>> I am using the semiisotropic pressure scaling because i want the box size
>>> to remain the same as that of the original box size in X and Y axis and want
>>> to change it only on the Z axis. I am doing this because at a latter stage i
>>> want to create an interface with organic solvent where i need the cross
>>> sections (X and Y axis length) of both the box should be the same.
>>>
>>> what i should do now get the pressure of 1 bar?
>>>
>>> 7 +/- 500 *is* approximately 1 bar, and hardly any better an approximation
>>> than 1 +/- 500. If you want lower fluctuations, use a larger system and run
>>> for a much longer time. Or, since you'll have to re-equilibrate once you
>>> combine the solvent boxes, don't bother.
>>>
>>> Mark
>>>
>>> Has no one has got the pressure of water to be close to 1 bar in GROMACS
>>> till now?
>>>
>>> Regards
>>> Vinoth
>>>
>>> On Fri, Oct 29, 2010 at 12:54 PM, David van der Spoel
>>> <spoel at xray.bmc.uu.se> wrote:
>>>> On 2010-10-29 09.08, vinothkumar mohanakrishnan wrote:
>>>>> Hi Mark
>>>>>
>>>>> I read the link before posting the question. even though the fluctation
>>>>> is between 500-600 (as said in the link) bar the average pressure is
>>>>> around 7.4 bar. my concern is the average pressure?. Because at a latter
>>>>> stage i am going to combine this water box with another organic solvent
>>>>> and i want to avoid any complications there.
>>>> Think again of what you just wrote. Your value is 7 +/- 500. In fact
>>>> according to normal statistical rules you should round the value of 7 to 0.
>>>> If you want more accurate number you could increase the box size by a factor
>>>> of 100.
>>>>
>>>> By the way, why are you using semiisotropic pressure scaling in a water
>>>> box?
>>>>
>>>>> Regards
>>>>> Vinoth
>>>>>
>>>>> On Fri, Oct 29, 2010 at 12:22 PM, Mark Abraham<Mark.Abraham at anu.edu.au
>>>>> <mailto:Mark.Abraham at anu.edu.au>> wrote:
>>>>>
>>>>> On 29/10/2010 5:48 PM, vinothkumar mohanakrishnan wrote:
>>>>>
>>>>> Hi Gromacians
>>>>>
>>>>> I want to do equilibration of water (spc model) first in the NVT
>>>>> ensemble and then in the NPT ensemble to maintain a temperature
>>>>> of 300K and a pressure of 1 bar respectively. The NVT
>>>>> equilibration works fine and the average temperature turns out
>>>>> to be 299.229 K.
>>>>>
>>>>> Energy Average RMSD Fluct.
>>>>> Drift Tot-Drift
>>>>> ------------------------------
>>>>> -------------------------------------------------
>>>>> Temperature 299.229 10.3092 10.2463
>>>>> 0.0393873 3.93877
>>>>> Heat Capacity Cv: 12.494 J/mol K (factor = 0.00118698)
>>>>>
>>>>> when i do NPT equilibration i am not getting the desired
>>>>> pressure as 1 bar or atleast close to 1 bar (between 1-1.4 bar).
>>>>> In the mdp file i used semiisotropic pressure coupling type
>>>>> because i want to fix the length of the box same on two axis as
>>>>> that of the original box size and want to change it only on one
>>>>> axis. can any one tell me why iam not getting the desired
>>>>> pressure of 1 bar?.
>>>>>
>>>>>
>>>>> This looks normal for a smallish water system over 100ps. See
>>>>> http://www.gromacs.org/Documentation/Terminology/Pressure
>>>>>
>>>>> Mark
>>>>>
>>>>>
>>>>>
>>>>> Energy Average RMSD Fluct.
>>>>> Drift Tot-Drift
>>>>>
>>>>> -------------------------------------------------------------------------------
>>>>> Pressure (bar) 7.4339 574.052 573.574
>>>>> 0.812129 81.2138
>>>>>
>>>>> Given below is my mdp file (NPT equilibration). any help is
>>>>> highly appreciated.
>>>>>
>>>>> title = DCE NVT equilibration
>>>>> cpp = usr/bin/cpp
>>>>> integrator = md
>>>>> nsteps = 100000
>>>>> dt = 0.001
>>>>> nstxout = 100
>>>>> nstvout = 100
>>>>> nstenergy = 100
>>>>> nstlog = 100
>>>>> ns_type = grid
>>>>> nstlist = 1
>>>>> rlist = 1.0
>>>>> coulombtype = PME
>>>>> rcoulomb = 1.0
>>>>> vdwtype = Cut-off
>>>>> rvdw = 1.0
>>>>> pme_order = 4
>>>>> fourierspacing = 0.16
>>>>> pbc = xyz
>>>>> tcoupl = V-rescale
>>>>> tc-grps = system
>>>>> tau_t = 0.1
>>>>> ref_t = 300
>>>>> pcoupl = berendsen
>>>>> pcoupltype = semiisotropic
>>>>> tau_p = 0.5
>>>>> ref_p = 1.0 1.0
>>>>> compressibility = 0.0 4.5e-5
>>>>> DispCorr = Enerpres
>>>>> gen_vel = yes
>>>>> gen_temp = 300
>>>>> gen_seed = 173529
>>>>>
>>>>> Regards
>>>>> Vinoth
>>>>>
>>>>>
>>>>> --
>>>>> gmx-users mailing list gmx-users at gromacs.org
>>>>> <mailto: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
>>>>> <mailto:gmx-users-request at gromacs.org>.
>>>>> Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>>>
>>>>>
>>>>
>>>> --
>>>> David van der Spoel, Ph.D., Professor of Biology
>>>> Dept. of Cell& Molec. Biol., Uppsala University.
>>>> Box 596, 75124 Uppsala, Sweden. Phone: +46184714205.
>>>> spoel at xray.bmc.uu.se http://folding.bmc.uu.se
>>>> --
>>>> 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