Coulomb energies and forces incorrect with coulomb-type = pme-user
Under group cutoff-scheme, simulation temperature only accords with the ref_t in nh-chain-length=1 in nose-hoover thermostat.
Using nh-chain-length!=1 of nose-hoover or v-rescale thermostat, I can't get the simulation temperature according with ref_t and the simulation temperature is higher than ref_t.
My system uses the tutorial materials from Lysozyme in Water
Topol file, mdp file and system after EM are attached.
The amber99sb-ildn of nbfunc=1 and comb-rule=1 version is here
The command I used:
gmx_mpi grompp -f nvt_group.mdp -c em.gro -p topol.top -o nvt_99sbildn_functype11_user.tpr
mpirun -np 20 gmx_mpi mdrun -quiet -ntomp 1 -nb cpu -notunepme -deffnm nvt_99sbildn_functype11_user
#2 Updated by Berk Hess about 1 year ago
- Subject changed from Group cutoff-scheme and simulation temperature to Coulomb energies and forces incorrect with coulomb-type = pme-user
- Status changed from New to Fix uploaded
- Priority changed from Normal to High
- Target version set to 2016.5
- Affected version changed from 2016.3 to 2016
This is not a temperature coupling issue, but a bug in the pme-user interaction implementation. I uploaded a fix for this.
Release 5.1 and before do not seem to be affected by this.
PS You should never change the combination rule of a force field. The warning in grompp is given for the case a user makes his/her own force field.