[gmx-users] COM motion removal: diffusion analysis

[Justin Lemkul]

On 12/28/15 10:28 AM, gromacs query wrote:
> Hi All
>
>
> I have protein-membrane system and gromacs tutorial by Justin suggests to
> do COM separately for protein_membrane and water. I am bit meesed up COM
> removal understanding.
>
>
> 1) I want to do diffusion analysis of protein in a membrane (2D/3D) how it
> will influence my results if I perform it on a trajectory with COM removal
> as seperate groups or on trajectory with COM removal of system.
>

The reason that interfacial systems are treated this way is because the 
diffusion of the two interfaces is different, hence must be treated separately.

> 2) Is it possible to remove COM seperately on a trajectory created using
> COM removal on system. e.g. for msd calculation one can use -rmcomm
>

IIRC the use of -rmcomm should have no effect if COM motion removal was already 
applied during the simulation.

> 3) May be am wrong here: In case of simple peptide in water box, and I
> remove COM for peptide and water (NOT as whole system) then in this case
> peptide should not cross the unit box. Is it right?
>

For a simple solute in solvent, you should not separate groups (I have even seen 
this result in crashes before).  COM motion removal is used to counteract 
spurious accumulation of translational kinetic energy as a result of using 
thermostats.  Read about the old "flying ice cube" problem.  It is not a matter 
of resetting positions; it is about subtracting falsely accumulated kinetic energy.

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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