VMD-L Mailing List

From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Thu Jun 09 2005 - 14:02:18 CDT

There's one caveat with this approach. CHARMM and psfgen won't always put
atoms in the same order in the psf file. NAMD and VMD always assume that
the psf, pdb, and dcd files have the same atom ordering. (The psfgen
coordpdb command is the only exception to this.) If you generate a
trajectory file using a psf file from psfgen, and then try to read the
trajectory using a new psf file generated in CHARMM this may bite you.
This is a general property of trajectory files since most formats don't
include atom name information.

If you want to use CHARMM for analysis then you might have to use CHARMM
to generate the psf file in the first place. Translating trajectories
etc. between different toolchains is an unsolved problem, AFAIK.

-Jim

On Wed, 8 Jun 2005, Chang, Christopher wrote:

> Leonardo,
>
> I ran into a similar problem when trying to move from NAMD to
> CHARMM--the "CHARMM format" PSF file that psfgen generates isn't
> apparently in CHARMM format. I rebuilt the molecules I was working on
> from scratch in CHARMM. You should be able to do that, then use whatever
> coordinate file you want--the PSF file doesn't contain coordinate
> information (just connectivity).
> So, if you already have a NAMD trajectory, rebuild the system in
> CHARMM using sequence files and the output PDB file from your
> simulation. Depending on the number of segments in your system (protein,
> solvent, ions, etc.), you may have to parse the single PDB file you have
> into multiple files, but that shouldn't be hard. With luck, the internal
> coordinates CHARMM generates will be the same ones that VMD generated;
> if not, you could edit them in CHARMM, or manually reformat the
> appropriate sections of your NAMD PSF file and paste them into the
> CHARMM PSF file.
>
> Chris
>
> Christopher H. Chang, Ph.D.
> Research Associate
> National Renewable Energy Laboratory
> 1617 Cole Blvd., Mail Stop 1608
> Golden, CO 80401
> Phone (303) 275-3751
> Fax (303) 275-4007
>
>
>
>> -----Original Message-----
>> From: owner-vmd-l_at_ks.uiuc.edu
>> [mailto:owner-vmd-l_at_ks.uiuc.edu]On Behalf
>> Of Leonardo Sepulveda Durán
>> Sent: Wednesday, June 08, 2005 11:55 AM
>> To: vmd-l_at_ks.uiuc.edu
>> Subject: vmd-l: Problems with Psf using Charmm format, solvate and
>> autoionize
>>
>>
>> Hello to everyone
>>
>> I am new in NAMD, and I am trying to use CHARMM program to analyze my
>> NAMD dcd trayectories, as in the manual is said to be posible. I found
>> this thread in VMD site:
>>
>> http://www.ks.uiuc.edu/Research/vmd/mailing_list/vmd-l/3561.html
>>
>> which talks about the dificulties of translating X-plor Psf files to
>> Charmm ones or vice-versa.
>> Most of analysis programs of CHARMM requires psf files to be in CHARMM
>> format (I have not found a comand to allow x-plor formated psfs). In
>> the thread above is said the psfgen command "writepsf" can write
>> x-plor and charmm psf files using
>>
>> writepsf [xplor|charmm] file.psf
>>
>> so I modify a vmd tcl script I wrote with lines stolen from various
>> other scripts, which use to solvate and ionize my protein:
>>
>> ##############################################
>> 1 mol load pdb 1UBQ.pdb
>> 2
>> 3 set ubq [atomselect top protein]
>> 4 $ubq writepdb ubq-p.pdb
>> 5
>> 6 package require psfgen
>> 7 topology /home/RPMs/NAMD/toppar/top_all22_prot.inp
>> 8 pdbalias residue HIS HSE
>> 9 pdbalias atom ILE CD1 CD
>> 10 segment PROT {pdb ubq-p.pdb}
>> 11 coordpdb ubq-p.pdb PROT
>> 12 guesscoord
>> 13 writepdb ubq.pdb
>> 15 writepsf charmm ubq.psf
>> 16
>> 17 package require solvate
>> 18 solvate ubq.psf ubq.pdb -t 5 -o ubq-wb
>> 19
>> 20 mol load psf ubq-wb.psf
>> 21 mol load pdb ubq-wb.pdb
>> 22
>> 23 set everyone [atomselect top all]
>> 24 measure minmax $everyone
>> 25 measure center $everyone
>> 26
>> 27 package require autoionize
>> 28 autoionize -psf ubq-wb.psf -pdb ubq-wb.pdb -is 0.00 -o
>> ubq-wb-ionized
>> 29 exit
>> ############################################
>>
>> the script works well if the option "charmm" in line 15 is omited .
>> Nevertheless, if I use it, it crashes after line 18 with the following
>> message:
>>
>> PSF file is in CHARMM format; XPLOR format required.
>> error reading atoms
>>
>> MOLECULE DESTROYED BY FATAL ERROR! Use resetpsf to start over.
>>
>> I dont know, but maybe solvate cannot use charmm-format psfs. At least
>> I haven't found a way to make solvate work in that context. As
>> autoionize cannot be called, I dont now either whether it allows
>> charmm psfs. Taking this into account, I would like to ask two
>> questions:
>>
>> 1) It is posible to modify this script to obtain a charmm psf of my
>> protein solvated and ionized? I thought it may be posible using a
>> modified solvate and autoionize routine, but I dont know if someone
>> have done this before, or where ca i find them.
>>
>> 2) Are dcd files psf-format independent, so a dcd file created using a
>> x-plor psf can be read after modifiing the psf to charmm format? if
>> is posible, I would be able to analyze my x-plor psf-generated
>> trayectories in charmm, otherwise I'll have to do all again.
>>
>> any help or suggestion is greatly acknowledged
>>
>> Leonardo
>>
>>
>