[gmx-users] "Mutating" type BAR calculations

[jrustad]

Dear All

Although the tutorials on FE calculations are all excellent and useful, they
all appear to involve turning a molecule on or off.  As I understand the
manual, it is also possible to change one atom into another, i.e. lambda=1
means 100% state A (and 0% state B) and lambda=0 means 100% state B (and 0%
state A)

I would assume in this case that "couple-lambda0" and "couple-lambda1" are
both set to "vdw-q", in other words, the VDW and coulomb interactions are
both "on" for each state?

I say this because I am having strange behavior where the VDW terms for
state A and state B appear to be added together at lambda=1.  In other
words, the VDW terms for state A were not turned off at lambda=1.  This
looks like a bug to me, however, I am too new to GROMACS to be comfortable
coming to such a conclusion.

I should also mention that I am using user-defined tables for the
short-ranged interactions.  From the manual, I cannot tell for sure whether
such tables are definitely supported in BAR (or, I suppose, in Free Energy
calculations in general)

I would be grateful for any insight into this problem

Jim Rustad

--
View this message in context: http://gromacs.5086.n6.nabble.com/Mutating-type-BAR-calculations-tp4997768.html
Sent from the GROMACS Users Forum mailing list archive at Nabble.com.