[gmx-users] calculating energy between certain groups

[Justin Lemkul]

On 5/15/18 3:26 PM, Irem Altan wrote:
> Hi,
>
> I ran a simulation using GPUs, and now I would like to calculate the energy between two protein chains, as well as energy between the protein and the solvent. For this, I am trying to run enemat:
>
> gmx enemat -f umbrella0.edr
>
> but this fails:
>
> Opened umbrella0.edr as single precision energy file
> Will read groupnames from inputfile
> Read 2 groups
> group 0WARNING! could not find group Coul-SR:Protein-Protein (0,0)in energy file
> WARNING! could not find group LJ-SR:Protein-Protein (0,0)in energy file
> WARNING! could not find group Coul-SR:Protein-SOL (0,1)in energy file
> WARNING! could not find group LJ-SR:Protein-SOL (0,1)in energy file
> group 1WARNING! could not find group Coul-SR:SOL-SOL (1,1)in energy file
> WARNING! could not find group LJ-SR:SOL-SOL (1,1)in energy file
>
> I read here (https://redmine.gromacs.org/issues/1822) that this information is not written into the .edr files when the simulation is run on GPUs. I am using version 2016.3, but there are some simulation results with 5.1.2 that I would like to analyze as well. Is it true that this is not possible with my setup? If so, is there a workaround like calculating energies given individual snapshots, rather than having to deal with .edr files?

The requested energy decomposition is simply not performed on GPU, 
though older versions just printed out zero for all group combinations. 
Recalculate the energies with a CPU-only run via mdrun -rerun. Energy 
decomposition is, in any case, basically just analysis so computing the 
energies on the fly is neither necessary nor efficient.

-Justin

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

Justin A. Lemkul, Ph.D.
Assistant Professor
Virginia Tech Department of Biochemistry

303 Engel Hall
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.thelemkullab.com

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