[gmx-users] Simulations in extreme conditions

[Justin Lemkul]

On 2/15/14, 6:52 AM, CipPruteanu wrote:
> Justin Lemkul wrote
>> I'm assuming that NPT and the data collection period are effectively the
>> same,
>> since you likely don't have anything worth restraining to distinguish any
>> sort
>> of distinct equilibration period.  The main question is: do your systems
>> equilibrate at the desired conditions?  25 kbar is a bit crazy for the
>> biomolecular force fields in Gromacs, and given the inherent fluctuations
>> in
>> pressure coupling, I would be a bit surprised if you actually achieved
>> such
>> pressures in that amount of time.
>
> Dear Justin,
>
> Thank you very much for the prompt reply. I am aware that it is very unusual
> to try to use GROMACS for extreme conditions simulations (such as inside
> planets and satellites, which is where this system is found in reality).
>
>
> Far as I can tell (and also people who do this for biological systems on a
> daily basis, at our university) the system equilibrates properly. It reaches
> the desired pressure and temperature. The fluctuations are large but
> bearable on the scale of the system (less than 0.5 kbars for 25 kbars and
> less than 1 kbar for 150 kbars, from my trials so far). There doesn't seem
> to be a problem with excessive pressure scaling (only once GROMACS gave the
> warning that the pressure scaled too much during a NPT equilibration, and it
> turned out it was my fault since I forgot to do the NVT equilibration before
> the NPT one).
>

Well that's good news.  Barostats are easy to break, but apparently yours are 
behaving.

>
> Justin Lemkul wrote
>> The biggest culprit, in my mind, is the force field.  OPLS and other
>> biomolecular force fields weren't designed for extreme conditions like
>> this, and
>> interactions with water under ambient, biological conditions, will be
>> wildly
>> different from what you're trying to do.  That's not to say it can't be
>> done, of
>> course, but the "standard" parameters may no longer apply.
>>
>> -Justin
>
> I am completely aware of this, but the simulations raise some interesting
> problems even in this situation. They do successfully show some physical
> trends and even something resembling what we are expecting from experiment,
> to a certain degree. Some people seem to find it very interesting like this,
> and are quite keen to push things just to see what happens and maybe this
> would point to some specific shortcomings of the forcefield.
>

That's all I was really getting at.  Some things may work, others may not. 
Pushing the force field(s) to extremes is an interesting venture; hopefully it 
works out for you.

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA 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 | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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