[gmx-users] index file for interaction energies

Justin A. Lemkul jalemkul at vt.edu
Tue Feb 22 21:13:14 CET 2011

Nick wrote:
> Dear experts,
> 1- I am trying to get interaction energies between solute (3 chains A) 
> and solvent 50 molecules B. In the index file I created two groups: one 
> for all atoms of [A] and and [B] for all solvent molecules. and by 
> setting A and B as energy groups in mdp file I am getting break down as 
> A-A, A-B and B-B with g_energy.
> I am a little confused as I dont know if I need to create different 
> groups for my solute to get interaction energies A-A, A-B and B-B. I 
> mean do I need to have [A1] for solute chain 1 ...[A3] for chain 3?

If you want a breakdown of per-chain energetics, then yes, specify each chain as 
a group.  The programs will only do what you tell them, nothing more, nothing less.

> and then sent A1, A2, A3 in mdp file and run g_energy? if I need to 
> specify different chains, then to get say A-A should I average over all 
> possible A1-A2, A1A3, A2-A3, A1-B, A2,B...? I am really confused ..

Theoretically, the short-range terms should sum, not average.

> 2- I know it is naive question but What does A1-A1 mean? How can one 
> chain interact with itself?

Without knowing what the chain is, no one can answer this.  Generally, any atoms 
that are within the short-range cutoff and further away than nrexcl bonds 
contribute to short-range interactions.  Long-range interactions (i.e., PME 
terms) happen too, but you can't decompose that term with energygrps.

> 3- Do I need to use mdrun - rerun option to get beakdown or just 
> g_energy gives what I need?

You need to -rerun.  g_energy does not take an index file, and it only analyzes 
existing groups, it cannot derive new ones.


> Thanks for your help
> Paniz


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