[gmx-users] Free energy calculation problem with too high dG

[Vlad]

Dear GMX community

I am trying to perform free energy calculation and encounter a problem similar to that in the discusion in September (theme "Free Energy of Binding Question")
This problem arises both in the complex with protein and with the ligand in water. It appears in the step where ligand starts to recouple (I start from the decoupled ligand and consequently turn on vdw and then charges). And the dG between the zero and the first init_lambda appears about 1800 kJ/mol ( This problem remains if we use npt.xvg files for BAR instead of md). All other dG are not higher than 10 kJ/mol and all negative. But the first one is too high and positive.

Here are parameters concerning free energy calculations.

free_energy=yes 
couple-moltype = UNK
couple-intramol = no
couple-lambda0 = none
couple-lambda1 = vdw-q
init_lambda_state = 0
calc_lambda_neighbors = 1
vdw_lambdas = 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
coul_lambdas = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
nstdhdl = 100
Also please find attached  energy minimization, nvt and npt mdp files and files from the BAR analysis:
http://turbobit.net/6ow5sa65wn6x.html  

This problem remains when switching between Gromacs2016 and Gromacs 5.1.4.


Sincerely yours,
Vladislav Balaev