[gmx-users] a question about energygrps

Justin A. Lemkul jalemkul at vt.edu
Wed Apr 25 16:34:54 CEST 2012

On 4/25/12 10:20 AM, Albert wrote:
> hello Justin:
> thank you very much for kind reply.
> In gromacs tutorial I found that the author use the following paramters:
> tc-grps = Protein DPPC SOL_CL ; three coupling groups - more accurate
> comm-grps = Protein_DPPC SOL_CL

OK, just so we're clear: tc-grps, comm-grps, and energygrps are all wildly 
different concepts, so don't get them confused.

> do you have any idea why did he use different groups for above paramters in
> NVT.mdp?

Indeed I do - I wrote it ;)

Ideally, one would like to use a single thermostat for the whole system, but in 
practice, especially for heterogeneous systems, the heat exchange between 
different groups can be different.  Hence the protein, lipids, and aqueous 
solvent are coupled separately.  They are of sufficient size to justify their 
own group (note, for instance, that ions are not coupled separately).

With respect to the comm-grps, since the system is basically an interface, the 
lipids and aqueous layers can slide with respect to one another, generating no 
net COM motion, but each layer may actually be have a net velocity, which would 
lead to artifacts.


> thank you very much
> On 04/25/2012 04:13 PM, Justin A. Lemkul wrote:
>> On 4/25/12 10:07 AM, Albert wrote:
>>> Hello:
>>> I am running a membrane simulation with gromacs and I wondering how to deal with
>>> energygrps? Should I put protein and lipids into one energygrps? Or I should
>>> leave the lipids stay with solvent and ions?
>> You can divide the system in any way you like for energygrps; it's just a
>> decomposition of the nonbonded terms. The way your break it down depends on
>> what you care to measure in the system.
>> -Justin


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list