[gmx-users] Negative part of the reaction coordinate in PMF

[Alex]

Dear all,

In a PMF simulation using WHAM, the windows are somehow that the reaction
coordinate (RC = COM_group1 ----COM_group2) should have both positive and
negative parts along Z direction, e.g COM_group1 locates in the middle of
the reaction coordinate more or less. However, what the WHAM gives out is
only the positive part of the RC, while all the .tpr and pullf.xvg files
for the whole RC are available.

I wonder how I can have PMF along the whole reaction coordinate in this
case?
Of course, one possible solution might be to run WHAM separately on windows
of positive and negative parts of reaction coordinate and then merge the
two profiles in a single profile later. But what I found out unexpectedly
is that presence of each part (e.g negative part) in gmx WHAM, would affect
on the PMF of the another part so that the PMF of the only positive part
varies when the windows of the negative part are and are not considered.
The -sym yes flag doesn't help as the two end points of the RC don't
overlap each other.

Thank you,
Alex