[gmx-users] PMF calculation with constraint distance
jalemkul at vt.edu
Thu May 21 23:04:33 CEST 2015
On 5/21/15 5:00 PM, leila salimi wrote:
> Hi all,
> I started to calculate PMF between two ions in solution and I
> used constraint for pull option, instead of umbrella as I saw in Umbrella
> sampling Tutorial of Justin.
> This is the part of my mdp file!
> ; Pull code
> pull = constraint ; Justin's value was "umbrella"
> pull_geometry = distance ; simple distance increase
> pull_dim = Y Y Y
> pull_start = no ; Justin's value was "yes"
> pull_ngroups = 1
> pull_group0 = CA_surf
> pull_group1 = carboxylate_carbon
> pull_init1 = 0.288 ; initial distance between pulling groups
> pull_rate1 = 0.001 ; 0.001 nm per ps
> pull_k1 = 1000 ; kJ mol^-1 nm^-2
> I got the output and I tried to pull out with increasing the distance
> between two atoms and I have pullf.xvg file!
> I know that I can have the time evolution of constrained distance with
> g_dist from xtc file! I am wondering how to have the distance with
> corresponding constrained force?
> If I have the file contain forces and distances, I can integrate the force
> then I will have PMF.
This probably won't give you a very reliable PMF. You're effectively trying to
use the Jarzynski method (albeit with a constraint rather than a harmonic
potential) with one run (normally you have to do this many, many times), which
won't give a correct result.
> Could you help me that how I can manage this?
If you have the constraint force in pullf.xvg and the distance in another .xvg
file, it's just a matter of using grep/awk to combine the two. Whether or not
this means anything is another matter. Getting a PMF between ions is trivially
simple with a harmonic potential.
Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow
Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201
jalemkul at outerbanks.umaryland.edu | (410) 706-7441
More information about the gromacs.org_gmx-users