GROMACS version: 2924-2
GROMACS modification: No
I am running absolute FEP with a small ligand and protein ( solvation and binding ) and gromacs is performing well. Lambda’s were constructed so that vdw coupling appeared before coulomb. Results for reported del G values were reasonably close ( ± 10%) . Itp’s were taken from acpype and oplsaaff from gromacs were used. The ligand is ionized but in going from, say , annihilation to solvation there is an assumed charge imbalance since the counter-ions were not altered. So, I am curious, does the FEP calculation take ion charge into account ?
Out of curiosity, I included the counter-ions as part of the ligand in a test run. The del G for solvation changed from ~ -1000kJ/mol to ~300kJ/mol. I did not (yet) run without any ions since gromacs gets upset doing so.
Based on reported data, it appears I should leave the ions intact. Opinions ?
thanks
PB