GROMACS version: 2025.3
GROMACS modification: Yes/No
Here post your question:
Hello I CHARMM_GUI generated a gromacs step6.0_minimization.mdp file for a large transmembrane-ligand complex, the below error/notes were generated subsequent to Gromacs execution:
NOTE 1 [file step6.0_minimization.mdp]: With Verlet lists the optimal nstlist is >= 10, with GPUs >= 20. Note that with the Verlet scheme, nstlist has no effect on the accuracy of your simulation.
NOTE 2 [file step6.0_minimization.mdp]: You have set rlist larger than the interaction cut-off, but you also have verlet-buffer-tolerance > 0. Will set rlist using verlet-buffer-tolerance.
ERROR 1 [file step6.0_minimization.mdp]: With switched vdw forces or potentials, rvdw-switch must be < rvdw
The CHARMM_GUI generated minimization.mdp file used is exhibited below:
Is this CHARMM36m force field? In that case both rcoulomb and rvdw should be 1.2 nm, as you can see also in the manual. This is happening because you are telling GROMACS to apply a force-switch modifier in the interval rvdw_switch - rvdw, which in this case would be a zero-width interval and would probably leave you with a discontinuity of the forces at the boundary.
Hello obZehn thank you for your kind update:) and Google AI Mode indicates Yes, CHARMM-GUI Membrane Builder utilizes the CHARMM36m force field. And yes this was the CHARMM direct output. Accordingly I will modify as suggested, thanks again:)
CHARMM-GUI actually doesn’t support only CHARMM36m but they have also other FFs implemented, which you can chose when you are building your box. I never had this problem of getting the wrong vdw switch length in output, usually this is correctly set to 1.2 by CHARMM-GUI, that’s why I am asking. So double check the force field, but if it’s indeed CHARMM36m then go ahead with the parameters and electrostatics as listed in GROMACS manual! :)
Hello obZehn thank you for your kind update:) and yes I streamed through CHARMM-GUI similarly to confirm and it is set by default to CHARMM36m, accordingly I modified as firstly indicated and it was executed and running, thanks:)
