[gmx-users] Looking for advice on Monte Carlo simulations, please

[Andrew S. Paluch]

The only open source Monte Carlo package I have every used or now of is 
MCCS Towhee:

http://towhee.sourceforge.net/

Andrew

On 11/08/2013 11:23 AM, Andrew DeYoung wrote:
> Hi,
>
> I have been using Gromacs for my MD liquid simulations for about 2 years
> now, and, of course, it has been working great!
>
> Now, in the current part of my project, I am looking at liquid structure,
> not dynamics.  Of course, one can analyze liquid structure by simply
> analyzing an MD trajectory.  I have run into some difficulties with
> sampling, however.  I could try some "special" sampling techniques (umbrella
> sampling, replica exchange, etc.), but another thing that has been suggested
> to me is to try Monte Carlo simulations.  Monte Carlo simulations, of
> course, do not model dynamics, but since I'm only interested in structure at
> this point, Monte Carlo might be a faster, less computationally expensive
> way to do sampling of possible liquid structures/configurations.
>
> It looks like, according to the Gromacs website, that a Monte Carlo
> integrator is in the process of being implemented in Gromacs (maybe for
> version 5.0 or so?).  If that is possible and will be funded, that would be
> great, because I would just like to use my same force field, same parameters
> as I have for my MD simulations--but just now I want to use Monte Carlo to
> quickly generate (low-energy/high-probability) configurations.
>
> But, until Monte Carlo is implemented in Gromacs, do you have any
> suggestions of another package that I can use to do MC on my system?
> Perhaps there is some MC package out there that is somewhat "similar" to
> Gromacs, such that maybe I could write scripts to translate my system
> topology from Gromacs to the MC package?  Do you have any experience or
> advice on Monte Carlo simulations?
>
> By the way, I hesitate to even ask questions like "when is this or that
> feature going to be implemented in Gromacs?", because I know that the
> development of Gromacs is a huge project and that many people are spending
> lots of time and resources to continually add features!  I am not yet very
> fluent in C/C++ and am not much of a programmer.  I am more of a
> chemist/simulator.  So I just want to thank you all for all your work on
> making Gromacs what it is!
>
> Thanks!
>
> Andrew DeYoung
> Carnegie Mellon University
>

-- 

Andrew S. Paluch, PhD
Department of Chemical, Paper, and Biomedical Engineering
Miami University
PaluchAS at MiamiOH.edu
(513) 529-0784