[gmx-users] free energy calculation problem

[Jin Wen]

Hi all, 

I am using Gromacs4.5 for free energy calculations. At the beginning 6 lambda (0, 0.2, 0.4, 0.6, 0.8, 1.0) are used to test my system. 
There are two problems: 

1. dV/dlambda value at lambda=0.0 is more than 300,000 KJ/mol, which is ten times large than those at the other lambda values (around 30,000 KJ/mol). All simulations at each lambda value are done during 100 ps. Since all the dV/dlambda is extremely large, I am wondering what it is related to. 

2. Delta G value should be similar when the conformation is transfered from B state to A and in different sign. But integrations of all dV/dlambda are both positive in both cases (A to B, and B to A) in my results. 

I attach free energy stuff in the input file below. 

; Free energy control stuff 
free_energy              = yes 
init_lambda              = 0.100 
foreign_lambda           = 0.1 
couple-lambda0           = vdw-q 
couple-lambda1           = vdw 
couple-intramol          = yes 
nstdhdl                  = 10 
couple-moltype           = system 
delta_lambda             = 0 
sc_alpha                 = 0.5 
sc-power                 = 1.0 
sc-sigma                 = 0.3 


Thanks in advance, 

Jin