[gmx-users] Temperature Coupling and Coupling-Groups

[Justin A. Lemkul]

David van der Spoel wrote:
> On 3/1/10 5:13 PM, Justin A. Lemkul wrote:
>>
>>
>> shayamra at post.tau.ac.il wrote:
>>> Greetings dear Gromacs users,
>>>
>>> I would like to create a membrane-protein simulations (popc + protein,
>>> water, Na Cl ions, using Berendsen thermostat) and I am not sure which
>>> coupling-groups I should use.
>>> For example, why not couple the whole system together?
>>>
>>> I do realize that this is a large request on this type of mailing
>>> list. I will also be happy to receive references to relevant literature.
>>
>> Most reports in the literature describe protein, lipids, and
>> solvent+ions in separate groups. The reasons for not coupling the whole
>> system together are well-documented, and as a matter of fact, I just
>> updated the Gromacs wiki this weekend to point to several papers
>> describing this phenomenon:
>>
>> http://www.gromacs.org/Documentation/Terminology/Thermostats
> 
> That's interesting, thanks for putting this together Justin.
> I haven't read the paper from the Tavan group completely, but I don't 
> think they truly solve the problem, do they? It reads more like some 
> workaround.
> 

Right, I just included that paper because I came across it when exploring these 
thermostat issues myself, and I think they have a really nice discussion of some 
of the problems one faces when using thermostats, in general.  The paper also 
points to a lot of other references that delve more deeply into the underlying 
issues.

Hopefully these sorts of links will be useful.

-Justin

> The origin of the problem is the lack of coupling between the different 
> degrees of freedom, and as such one would expect united atom force 
> fields to show this effect to a larger degree than all-atom force 
> fields. It could also be that the use of constraints makes that the 
> effective number of degrees of freedom per atom is different for 
> proteins and water. It seems that this still hasn't been investigated 
> properly...
> 
>>
>> There are also numerous threads in the list archive on this topic,
>> likely with other references that I did not link.
> 
> 
>>
>> -Justin
>>
>>>
>>> Thanks in advance,
>>> -Shay
>>>
>>
> 
> 

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

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

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