[gmx-users] CG Force and Potential through do_force()

[Mark Abraham]

On 26/03/2010 3:26 AM, Sikandar Mashayak wrote:
> @ Mark :: you said if the potential is never used then I dont have to
> worry about storing its value. Suppose I am replacing Coulombic
> interactions with something of its equivalent through new force field,
> and I dont store the potential due to it, will I still get the similar
> results in terms of structure (RDF). Of course I am assuming that new
> force can accurately replace the Coulombic interactions among the
> charged atoms.
>
> By just having correct forces acting on the atoms and not worrying about
> storing potential energies, will I still get same trajectory as with
> storing potential energies.

Yes, if the potential is never used. In principle, MD is integrating 
Newton's Laws, which generally don't require the energy to be known. 
Whether that is true depends on details of your simulation protocol that 
we don't know. REMD, being a Monte Carlo method, needs an accurate 
potential. Free-energy methods might need an accurate potential. 
Generally speaking, equilibrium MD doesn't need it, however. You can 
roughly test for this by adding some arbitrarily large value to the 
energy (e.g. 10^90) and see if the subsequent MD breaks. First, do your 
homework about how the algorithms you're using work... :-)

Mark

> On Thu, Mar 25, 2010 at 5:45 AM, XAvier Periole <x.periole at rug.nl
> <mailto:x.periole at rug.nl>> wrote:
>
>
>     If your force is between "atom" pairs you might be better trying to
>     give all
>     of that as a tabulated potential ... no coding.
>
>
>     On Mar 25, 2010, at 7:14 AM, Mark Abraham wrote:
>
>
>
>         ----- Original Message -----
>         From: Sikandar Mashayak <symashayak at gmail.com
>         <mailto:symashayak at gmail.com>>
>         Date: Thursday, March 25, 2010 15:12
>         Subject: Re: [gmx-users] CG Force and Potential through do_force()
>         To: Discussion list for GROMACS users <gmx-users at gromacs.org
>         <mailto:gmx-users at gromacs.org>>
>
>             well if I need to add it by myself, I am not very sure in
>             what variable to add it to. Like I said, I know to add the
>             force on the atom i do f[indxofatom][XX]= forcevalue, where
>             to potential value due to this force?
>
>
>         It's not easy. Look at how the generic kernel does it and work
>         you way back through the data structures.
>
>         Mark
>
>             On Wed, Mar 24, 2010 at 10:52 PM, Mark Abraham
>             <Mark.Abraham at anu.edu.au <mailto:Mark.Abraham at anu.edu.au>>
>             wrote:
>
>
>             On 25/03/2010 2:42 PM, Sikandar Mashayak wrote:
>
>
>
>         Hi
>
>
>
>         I want to do Coarse-Grained simulation in Gromacs. I have one CG
>         force
>
>
>         formula, which I plan to implement by adding a short code in
>         do_force()
>
>
>         routine in Gromacs 4.0.7 source code.
>
>
>
>         After browsing through the do_force() code, I am now familiar
>         with data
>
>
>         structure relevant to add the new force. I mean, I can specify
>         new force
>
>
>         by adding to f[indx][XX] and so on.
>
>
>
>         But I am not very sure about potential, whether do I have to
>         explicitly
>
>
>         specify the potential on atoms due to this force or gromacs
>
>
>         automatically takes care of it by knowing the forces on each atoms?
>
>
>
>
>         You will need to add that yourself if you want the reported
>         potential to be accurate. GROMACS uses the given parameters in
>         its inner loops to calculate energies and forces. The former are
>         not reverse-engineered from the latter. If the potential is
>         never used, however, then you can forget about it...
>
>
>
>
>         Mark
>
>
>         --
>
>
>         gmx-users mailing list gmx-users at gromacs.org
>         <mailto:gmx-users at gromacs.org>
>
>             http://lists.gromacs.org/mailman/listinfo/gmx-users
>
>         Please search the archive at http://www.gromacs.org/search
>         before posting!
>
>
>         Please don't post (un)subscribe requests to the list. Use the
>         www interface or send it to gmx-users-request at gromacs.org
>         <mailto:gmx-users-request at gromacs.org>.
>
>
>         Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>
>
>
>             --
>             gmx-users mailing list gmx-users at gromacs.org
>             <mailto:gmx-users at gromacs.org>
>             http://lists.gromacs.org/mailman/listinfo/gmx-users
>             Please search the archive at http://www.gromacs.org/search
>             before posting!
>             Please don't post (un)subscribe requests to the list. Use the
>             www interface or send it to gmx-users-request at gromacs.org
>             <mailto:gmx-users-request at gromacs.org>.
>             Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>
>         --
>         gmx-users mailing list gmx-users at gromacs.org
>         <mailto:gmx-users at gromacs.org>
>         http://lists.gromacs.org/mailman/listinfo/gmx-users
>         Please search the archive at http://www.gromacs.org/search
>         before posting!
>         Please don't post (un)subscribe requests to the list. Use the
>         www interface or send it to gmx-users-request at gromacs.org
>         <mailto:gmx-users-request at gromacs.org>.
>         Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>
>
>     --
>     gmx-users mailing list gmx-users at gromacs.org
>     <mailto:gmx-users at gromacs.org>
>     http://lists.gromacs.org/mailman/listinfo/gmx-users
>     Please search the archive at http://www.gromacs.org/search before
>     posting!
>     Please don't post (un)subscribe requests to the list. Use thewww
>     interface or send it to gmx-users-request at gromacs.org
>     <mailto:gmx-users-request at gromacs.org>.
>
>     Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>
>