As you can see in previous results, the Group cutoff scheme has the same energy as NAMD/LAMMPS for Arg dipeptide, which means even inside Gromacs, the energies from the Group scheme and from the Verlet scheme are different. Is it normal that the electrostatic energies from these two cutoff schemes are different only for the charged dipeptides?
In my opinion, the energies for a simple dipeptide should be same no matter what MD programs or cutoff schemes we used, if all bonded and nonbonded interactions are included in the calculation. The weird thing is that the energies of neutral dipeptides with Verlet scheme are same with other programs, but the energies of charged dipeptdies are not, and the differences are all around 13.8 kcal/mol. However, the Group scheme with the same cutoff can provide the same energy for the charged dipeptides. That’s why I doubt that there might be some systematic errors in Verlet scheme when calculating energies for charged dipeptides.
I want the energies are consistent between different MD programs because it is important for library development. If a library is based on CHARMM force field energy, it cannot be generalized to all MD programs if the MM energy is not consistent in them. If you don’t want to dig into the Verlet codes, could you tell me what parameters about Verlet scheme I can set in mdp input file to obtain the same energy as the Group scheme, since the Group scheme is deleted in the current Gromacs version?