[gmx-users] problems with g_wham

[Gavin Melaugh]

Hi Guys

I am having great difficulty generating a histogram of the correct shape
from g_wham. I am running umbrella sampling on one configuration fo two
molecules at a fixed distance using the following
pull parameters. ( I am just running one configuration to begin with so
that can make sure that the shape of my distribution is correct).

pull        = umbrella
pull_geometry = distance
pull_dim = Y Y Y
pull_start = yes
pull_ngroups = 1
pull_group0 = cage_1
pull_group1 = cage_2
pull_init1 = 0
pull_rate1 = 0.0
pull_k1 = 2000
pull_nstxout = 100
pull_nstfout = 100

I have set the force constant quite high in the hope that the system
will sample the equilibrium distance much more often than the distances
accessible form the harmonic potential. In the plot of count vs distance
I get an upside down bell curve in which there is a larger distribution
at the extremeties of the curve. Does anyone know why this may be
happening? If so please let me know. Am I right in assuming that the x
axis in the histo.xvg file is the distance between the centres of masses
of the two molecules?

Cheers

Gavin