[gmx-users] Re: gmx 4.6.1, Expanded ensemble: weird balancing factors

[Michael Shirts]

Unfortunately, the semiiso fix for MTTK is not trivial, and will have
to wait until 5.0.

Berendsen volume + expanded ensemble would likely give very bad
results, especially since surface area fluctuations are important.

Replica exchange works with sd and parrinello-Rahman, if that helps!

I am working on checking various combinations of expanded ensemble and
integrators that can more easily be supported, but that may take a few
weeks.







On Wed, Jul 17, 2013 at 6:22 PM, Dejun Lin <dejun.lin at gmail.com> wrote:
> The test I did was using gmx-4.6.3. I'm currently working on a membrane
> system, which requires semiisotropic pressure coupling. I know that MTTK in
> gmx-4.6.3 doesn't support semiiso yet so the only combination available for
> expanded ensemble is md-vv + v-rescale tcoupl + berendsen pcoupl, which I'm
> testing right now. A preliminary run gave me some sane dG values although
> it's not optimal for data collection as you pointed out. I wonder if it's a
> trivial fix that might have been done to add semiiso to MTTK?
>
> Thanks again for your help!
>
>
> 2013/7/17 Michael Shirts <mrshirts at gmail.com>
>
>> > It seems to have something to do with the integrator/pressure-coupling.
>>
>> that is what I expected based on some preliminary testing earlier.
>>
>> When
>> > I ran the tutorial on
>> >
>> http://www.alchemistry.org/wiki/GROMACS_4.6_example:_Ethanol_solvation_with_expanded_ensemble
>> ,
>> > everything seems fine
>>
>> OK good!
>>
>> > unless I switched to sd integrator instead of md-vv
>>
>> There may be a problem with sd and expanded ensemble. This MAY have
>> been fixed in 4.6.3, but I'll need a little bit of time to check.
>>
>> > (and Pcoupl from MTTK to berendsen),
>>
>> Berendsen is provably wrong. The ensemble is incorrect.  It is evil
>> whenever a true distribution is needed. There are warnings that should
>> be printed when you set up the system.  Perhaps they should be made
>> stronger with expanded ensemble!
>>
>> in which case SHAKE didn't seem to be
>> > able to settle the constraints.
>>
>> Expanded ensemble is rough on constraints.  The integration is
>> formally correct, but it is harder to converge pressure, especially
>> with constraints.  The larger the system, however, the more stable it
>> is.
>>
>> > file from the tutorial. Any idea?
>>
>> Well, one solution is to run your molecule with the parameters in the
>> .mdp!  Would that solve your problem for now?
>>
>> Longer term. I will check on the SD to see if it is fixed in 4.6.3 --
>> I seem to re call this, but I'm not sure.   However, I am not sure
>> that it can be made formally correct.  I will either fix this or make
>> the warnings more clear -- or not permit it at all.
>> --
>> 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