[gmx-users] PMF in vacuum

[chris.neale at utoronto.ca]

Dear Eudes:

Can you please elaborate on your description? What .mdp options are  
you using? What exactly do your curves look like (you can post  
pictures to photobucket or some other online service and link them  
here)? If you suspect that you are doing something wrong, then we need  
to understand exactly what you are doing and exactly what you are  
seeing in order to help.

Chris.

-- original message --

Hi gmx-users,
I'm trying to simulate a umbrella sampling PMF between two benzene
molecules
in vacuum. My protocol is working fine, my histograms have good overlap and
the curves I have got are quite reasonable.

However I have noticed that some options in my .mdp file can significantly
change
the depth of the well. Also the curves can not go to zero at long distances.
For example, if I use PBC I get a reasonably good value for the minimum of
the PMF
but from a certain separation it starts to increase slightly in a linear
fashion
instead of going to zero. On the other hand, if I make pbc = no, I get an
acceptable curve,
with the PMF going to zero, but with the minimum too high.

Someone could give me any tips on the best set of parameters to calculate
this PMF in a vacuum?
bests
eef