[gmx-users] Free Energy Calculations in Parallel

[Justin Lemkul]

On 7/26/13 4:06 PM, Quintin Sheridan wrote:
> Dear Gromacs Users,
>
> Is it possible to run free energy calculations in parallel using mpirun? If

Yes, the free energy code is parallelized, either via MPI or threads.

> not I am wondering what the fastest way to run a free energy calculation
> is. I am basing my free energy calculation on the tutorial by Justin Lemkul
> found at
>
> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/free_energy/01_theory.html
>
> I am trying to use the Bennet Acceptance Ration method (g_bar) to get the
> free energy of solvation for an ionic liquid.  I have tried decoupling an
> ion pair from the system as well as individual ions.  These simulations
> will run locally but when I try to run them in parallel I get the error:
>
> Fatal error:
> There is no domain decomposition for 8 nodes that is compatible with the
> given box and a minimum cell size of 2.26125 nm
>

Consult 
http://www.gromacs.org/Documentation/Errors#There_is_no_domain_decomposition_for_n_nodes_that_is_compatible_with_the_given_box_and_a_minimum_cell_size_of_x_nm. 
  This topic is discussed frequently, particularly in the context of free energy 
simulations, because there are unique factors that come into play in terms of 
the DD algorithm.  Search the archive and you will surely find the detailed 
explanations.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441

==================================================