[gmx-users] Umbrella sampling - decomposing of free energy

[gozde ergin]

Dear gromacs users,

I did umbrella sampling in order to estimate the PMF profile of my system.
My system has 512 water molecules covered by 25 decanol molecules on both
side.
I pulled a water molecule bulk water to organic layer and than gas phase.
(Please see the system pic , http://imgur.com/sFFlVjF )

As a next step I wanted decompose this free energy in enthalpic and
entropic contribution
Delta G = Delta H - T Delta S

Enthalpies ΔH were computed by averaging the potential energy over the
respective simulation trajectory in each window (see the pic
http://imgur.com/tc3cO95 )

My question is as you see there is big variation in enthalpy.
What should I do to get more precise (less varied) potential energy?

Simulation time 4 ns, time step is 2 fs, force field os GAFF (produced by
topolbuild)
As an example here one of the window potential energy
http://imgur.com/Bw9aUpu

Thanks in advance