[14-XI-2012]
Hi,
Last friday, another colleague and me were discussing about computing the energy of a whole trajectory (using a different topology file than the original one) after it has been already generated (for example for FEP calculations).
He was aware of the option imin=5 but we currently do this process splitting the trajectory into individual snapshots, and then running a 1-step MD with the new topology file for every snapshot under the same conditions (mdin file) the trajectory was created. This process is really more tedious (and slower!!!) than just post processing the trajectory with imin=5 so I was thinking about the possibility of doing it according to the manual.
The reason to do this 1-step MD, instead of using the standard method was that he said that part of the subroutines in the minimization code did not compute the non-bonded, electrostatic, and Particle Mesh Ewald (PME) interactions in the same way than the MD subroutines (there was an old mail in the amber archives about this topic) so, if we just minimize the structure we will be evaluating energies incorrectly.
There is still the point about the minimization itself: is the energy after the 1-step minimization the energy of the input snapshot or the energy of the conformation AFTER the 1-step minimization has been done? In case that the second option is the right one we will be computing a lower value in energy for the whole ensemble of conformations and that sounds quite "disturbing" from the statistical mechanics point of view.
However, amber manuals 11 and 12 still suggest to use this method (imin=5) and say that for NVT calculations sander takes the dimensions of the box from the input file, precisely, to compute correctly the PME interactions.
Unfortunately, I am not yet an expert in the code so, my question is, would it be any problem to compute the total energy (and the single contributions: BOND, ANGLE, DIHED, VDW, EELEC, 1-4EL, 1-4NB) by using the imin=5 option rather than splitting the trajectory and running 1-step MD for every single snapshot?
Thank you very much for your help.
Kepa K. Burusco
********************************************************************************
________________________________
De: Aron Broom <broomsday.gmail.com>
Para: AMBER Mailing List <amber.ambermd.org>
Enviado: Miércoles 7 de noviembre de 2012 6:24
Asunto: Re: [AMBER] Energy Analysis from Trajectory
oh wow, thanks. I really did look over the manual before asking that, but
as it's quite clearly stated, I have no excuses. Thanks for the reply :)
On Wed, Nov 7, 2012 at 1:09 AM, Roitberg,Adrian E <roitberg.ufl.edu> wrote:
> look in the manual for the imin=5 option in main
>
> that should compute the energy and components for every snapshot
> ________________________________________
> From: Aron Broom [broomsday.gmail.com]
> Sent: Wednesday, November 07, 2012 1:07 AM
> To: AMBER Mailing List
> Subject: [AMBER] Energy Analysis from Trajectory
>
> Hello AMBER users,
>
> If I have a trajectory and related parameter/topology files, and want to
> grab the potential energy for each snapshot in that trajectory (perhaps
> broken into various components, i.e. EEL and such), is the best option to
> use MM/PBSA.py? That is, even if I have no plan to break things down into
> receptor and ligand (I just have a single molecule) is it the simplest way
> to go about doing what I'm after, or will it be upset about not being able
> to identify a separate receptor and ligand?
>
> Thanks,
>
> ~Aron
>
>
>
> --
> Aron Broom M.Sc
> PhD Student
> Department of Chemistry
> University of Waterloo
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Aron Broom M.Sc
PhD Student
Department of Chemistry
University of Waterloo
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 14 2012 - 14:00:03 PST
