running LAMMPS with CMAP for system modeled with multiple CHARMM force fields

Dear all,

I am trying to run a simulation of a polymer-protein system in LAMMPS modeled with CHARMMC35r, CGENFF and CHARMMC36 force fields. I was able to successfully convert the files into LAMMPS format using the script (CMAP version) by cut/paste of respective parameter files into one. However, I run into an error when I try to run the MD simulation. When I broke down my script, I found that the following MPI error pops up at the step when I am reading the LAMMPS data file. (Please find the files in the OneDrive link:


units real
neigh_modify delay 2 every 1

atom_style full
bond_style harmonic
angle_style charmm
dihedral_style charmm
improper_style harmonic

pair_style lj/charmm/coul/long 10 12
pair_modify mix arithmetic
kspace_style pppm 1e-5

Modify following line to point to the desired CMAP file

fix cmap all cmap charmm36.cmap
fix_modify cmap energy yes
read_data fix cmap crossterm CMAP

you have a broken geometry, please see the attached snapshot, which was created by reading the data file into VMD after commenting out the CMAP section. there are several “long” bonds that seem incorrect and cannot be explained by being a periodic continuation and thus misplaced image flags. with the most recent LAMMPS version, i get a more specific error message, which is cause exactly by such long bonds:

Finding 1-2 1-3 1-4 neighbors …
special bond factors lj: 0 0 0
special bond factors coul: 0 0 0
12 = max # of 1-2 neighbors
72 = max # of 1-3 neighbors
ERROR on proc 0: 1-3 bond count is inconsistent (src/special.cpp:292)
Last command: read_data fix cmap crossterm CMAP
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0

you will have to review and re-check the process of how you have built your geometry.

Dear Axel,

Thank you very much for your prompt response and suggestion. It was very helpful. I am reviewing and fixing the geometry building of the systems simulated.