Wrong data file gen using charmm2lammps.pl

Hi all,

I am trying to generate a lammps data file using charmm2lammps.pl script provided with lammps source code for alanine dipeptide.
I am using PDB file from NAMD tutorials as reference (attached adp.pdb). First, I generated a PSF file using psfgen module of VMD (attached pgn script file) and charmm36 protein topology file. Then I am using the PDB and PSF files obtained (attached adp_new.pdb, adp_new.psf) and charmm36 ff files to obtain my lammps data file (attached adp_new.data). My data file shows that there is just 1 atom type, which is not correct. Further information about about the topology of my system is missing from the data file. I fear my PSF file is wrongly generated, but I cannot figure out what is wrong with it. Also, the end of my PSF file shows there is 1 cross-term present, but the next line writes 2 quadruples (1 for each phi and psi). Ideally, there should be 2, as is written in the topology file as well.

It would be great if anyone can help me figure out what is going wrong with my psf and data files here.

Thanks,
Madhur Aggarwal

adp.pdb (1.77 KB)

adp_new.data (8.77 KB)

adp_new.pdb (1.75 KB)

adp_new.psf (4.83 KB)

psfgen_script.pgn (183 Bytes)

you didn’t say which command line you were using and what the output is. also you should let us know which LAMMPS version the charmm2lammps.pl script was taken from.
this is crucial information.

i have tried your files with the current development version of LAMMPS and i get a different, more reasonable looking data file. so i suspect you are using a version that has a bug or that you have used an unsuitable set of command line flags.

please note, that for a correct charmm36 conversion including CMAP correction, you need to correct the PDB file, as there will be an error with an explanation of what is missing and how to correct it.

so what you probably need to do is:

  • upgrade your LAMMPS version (or at least download the latest LAMMPS source package and use the charmm2lammps.pl script from there. also use the charmm36 files from the example-cmap folder).
  • correct your PDB file as needed so the inclusion of the CMAP correction in the conversion can work.
  • try the conversion again

axel.

Hi Axel,

Thanks for your valuable suggestions, I could generate the data file correctly from charmm2lammps.pl script available in latest 7th-Aug, 2019 version of lammps.

Madhur Aggarwal

LAMMPS users,

I noticed that the rich data-set by MacKerell Lab:

http://mackerell.umaryland.edu/charmm_ff.shtml

can be converted (with caution) by charmm2lammps.pl if the MASS index (-1 in the
new version, both topology and parameters files) is replaced as in the original version.

That is, in topology files:

MASS 1 H 1.00800 ! polar H
MASS 2 HC 1.00800 ! N-ter H

instead of:

MASS -1 H 1.00800 ! polar H
MASS -1 HC 1.00800 ! N-ter H

The index -1 drives charmm2lammps.pl into a big confusion.

Best,

Giovanni La Penna

National research council of Italy (CNR)
Institute for chemistry of organometallic compounds (ICCOM)
Sesto Fiorentino (Firenze), Italy
tel.: +39 0555225264
Published 2019: http://www.springer.com/978-3-319-95842-2