[gmx-users] Total Energy, g_energy

[Mark Abraham]

Hi,

On Sun, Nov 18, 2018 at 6:19 PM Vito GENNA <vito.genna at irbbarcelona.org>
wrote:

> Dear GMXusers,
>
> Here I am again with an unresolved problem.
> Briefly, in th active site of a system I've run, i have different
> molecules.
> As easily conceivable, I do expect different stabilities in terms of
> energy.


Note that you'll see different energies because they have different
composition, mass, available degrees of freedom, etc. One might observe a
stable total energy of 12 kj/mol for benzene in vacuo and 20 kj/mol for
cyclohexane in vacuo, but that doesn't mean anything about stability.

I am basically interested in getting something similar to Fig. 6 of
>
> https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3192800/figure/pcbi-1002189-g006/
> in order to highlight the different stabilities
>

Those graphs show that the conformations changed somewhat over the lifetime
of the simulation, but neither the timescale nor the amount of the change
is meaningful without an assessment that the starting or ending
configurations were of comparable quality. Nor can they tell whether e.g. a
rotamer inaccessible over the simulation time scale would be better or
worse.

What I did:
> by changing energygrps keyword in the md.mdp (*and maintaining System in
> the tc-grps*) I have generated 3 different .tpr
> 1) energygrps = AET PROTEIN
> 2) energygrps = BET PROTEIN
> 3) energygrps = CET PROTEIN
> At this point, I have used rerun command with the newly generated .tpr to
> get the new ener.edr files.
> So far it seems ok.
> Now I have two questions:
> 1) once i run g_enery I get all the possibile combinations of LJ and
> Coulombic terms of my energygrps. Great. However, I also get options:
> TOT.ENERGY, KINETIC, CONSERVED, POTENTIAL and so on.
> Are these referring to the groups declared in my mdp files or, differently,
> they refer to the whole system (thus comprising water, ions and so on).


That's the whole system. You can see that also from the fact that the
magnitude is consistent with that reported for the respective original
simulations.


> The
> question rises because if I measure the TOT.ENERGY of my systems they are
> almost overlapped!! (altought evidently one of the molecules has very bad
> interactions with the active site, and it is ready to depart)
>

If the kinetic energy dominates the magnitude of the potential energy
(whose absolute value was typically not parameterised to be meaningful), or
the ligand composition or pose are equivalent, then you will see about the
same total energy.


> 2) which terms (g_energy) should I use/combine to get a potential energy
> description of my molecules in my protein (Fig. 6 of the link I have
> reported)? I need some quick and qualitative (not FEP et similia)
>

They report total energy. Perhaps their ligands had a wider variety of
composition than yours. A quick glance at the text gives me no reason to
expect that their simulation results suggest even a qualitative conclusion.
That requires an assessment of how often conformations of particular
energies arise, bearing in mind that the total energy depends on the
composition, and so are not directly comparable between systems of
different composition.

Mark


> Thanks in advance for any kind contribution.
> I am stuck at this point!
>
> V
>
> --
>
> ***********************************************************************
> *Vito Genna, Ph.D*
> *EMBO Fellow*
>
> *Postoctoral Researcher*
>
> *Molecular Modeling and Bioinformatics*
> *Orozco Lab*
>
> *Institute for Research in Biomedicine (IRB Barcelona)*
>
> *Parc Centific de Barcelona*
>
> *C/ Baldiri Reixac 10
> <https://maps.google.com/?q=C/+Baldiri+Reixac+10&entry=gmail&source=g>-12*
> *08028 Barcelona*
>
> ***********************************************************************
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>