[gmx-users] Is "dVremain/dl" related to "fep-lambdas"?

[Igor Leontyev]

Dear gmx-users,
When I compute with gromacs-5.1 relative free energies using mixed 
topologies (i.e. by setting typeA, typeB, bond kA, kB ant etc) I see 
that some undocumented value "dVremain/dl" is printed in log- and xvg-files.

Please, clarify the exact meaning of the "dVremain/dl" quantity. How it 
is used in dG computation?



================== Some Thoughts ===================
Playing with different setup by setting parameters fep-lambdas, 
coul-lambdas, vdw-lambdas and bonded-lambdas, I figured out that
dVremain/dl of Setup-1 (when set only fep-lambdas) is equal to sum of 
all dV/dl components of Setup-2 (when all coul-lambdas, vdw-lambdas and 
bonded-lambdas are set explicitly):
dVremain/dl(1) = Vcoul/dl(2) + dVvdw/dl(2) + dVbonded/dl(2)

Setup-1)
# if in MDP-file
    ; coul-lambdas = 0.0 0.10
    ; vdw-lambdas = 0.0 0.10
    ; bonded-lambdas = 0.0 0.10
    fep-lambdas = 0.0 0.10
# then
    dVremain/dl
    1.52124e+03
#endif

Setup-2)
# if in MDP-file
    coul-lambdas = 0.0 0.10
    vdw-lambdas = 0.0 0.10
    bonded-lambdas = 0.0 0.10
    fep-lambdas = 0.0 0.10
# then
    dVremain/dl dVcoul/dl dVvdw/dl dVbonded/dl
    0.00000e+00 1.56939e+03 -6.97050e+01 2.15596e+01
#endif
=======================================================