Re: AMBER: generate charmm paramater file from amber parm file

From: David E. Konerding DSD staff <>
Date: Wed, 16 Feb 2005 10:13:23 -0800

HL Eastwood wrote:

>Dear amber users
>I have an amber derived pdb file and corresponding amber paramater file. Is it
>possible for me to generate a charmm parameter file from this data?
>An entry in the amber archive posted by Roass Walker suggests using ptraj:
>" would want the following input file for ptraj (for amber to charmm):
>trajin inputtrajectory.traj.gz
>trajout outputtrajectory charmm "
>I have done this and it generates a binary trajectory file for charmm. However,
>I still require the paramater file. Any suggestions on how to get this?

While you could write a program to extract the AMBER parameter data from
the prmtop file, and generate a CHARMm input file (after futzing with
all the values to make them more CHARMm-compatible), it would probably
be easier to minimally reformat the
PDB file so that CHARMm or NAMD's psfgen could convert it to a .psf
file. Unfortunately,
I think NAMD's psfgen has slightly different definitions for protein and
nucleic acid termiii,
and it sometimes adds atoms; you have to be very careful in your atom
and residue naming to ensure everything works the same (IE, you end up
with a PSF file that has a one to one mapping
to the coordinate identity and ordering as the prmtop and AMEBR coords)

Here's an example psfgen input file that I used with my PDB file(s)
starting from 1A31 or 1A35.pdb, after I did a bunch of reformatting:

topology top_all27_prot_na.inp
alias residue HIS HSD
alias atom ILE CD1 CD
alias atom PHE OXT OT1

segment topo {
    pdb topoI.pdb
coordpdb topoI.pdb topo

alias residue A ADE
alias residue G GUA
alias residue C CYT
alias residue T THY

segment dnaC {
    first none
    last none
    pdb dnaC.pdb
coordpdb dnaC.pdb dnaC

segment dnaD {
    first none
    last none
    pdb dnaD.pdb
coordpdb dnaD.pdb dnaD

writepdb topoI_dna.psf.pdb
writepsf topoI_dna.psf

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Wed Feb 16 2005 - 18:53:01 PST
Custom Search