Reading topology file

[Chris]

Hi,

I am trying to read into CP2K .pdb and .psf files for a system, but I
get the message:

*****************************************
*** ERROR in topology_molecules_check ***
*****************************************

*** different atom name for same molecule kind molecule type =
5 ***
*** molecule number= 2 expected atom name=OCL
found=CTL2 ***

I have checked my input .psf file, and I only have one type of
molecule containing atoms of type OCL. In the .psf file, all of the
occurrences of atoms with the name OCL occur in the same place
relative to the other atoms in the molecule they are in, and the atoms
in the .psf file are (so far as I can tell) in exactly the same order
as they are in the .pdb file.

Please could someone explain this error message?

Thanks,

Chris

[Axel]

On Dec 12, 1:34 pm, Chris <christopher.illingwo...@chem.ox.ac.uk>
wrote:
> Hi,

hi chris,

>
> I am trying to read into CP2K .pdb and .psf files for a system, but I
> get the message:
>
> *****************************************
> *** ERROR in topology_molecules_check ***
> *****************************************
>
> *** different atom name for same molecule kind molecule type =
> 5 ***
> *** molecule number= 2 expected atom name=OCL
> found=CTL2 ***


> I have checked my input .psf file, and I only have one type of

how did you generate your pdb/psf file. please keep in mind,
that you cannot expect that if your input works with one package
that it'll work with a different one just as well. many are
"broken" in very unique ways. it also matters as what you read
in files (there is UPSF and PSF for example).

however, it is also possible, that your compiler miscompiled the
code. i remember seeing similar error messages when i was on a
machine that had PGI 6.2.x compilers and those went away when they
were upgraded to 7.0.2. i'm now always compiling a g95 based
version on top of the system default compilers, so that if there
are input errors, i can do a cross check with an unrelated compiler.

> molecule containing atoms of type OCL. In the .psf file, all of the
> occurrences of atoms with the name OCL occur in the same place
> relative to the other atoms in the molecule they are in, and the atoms
> in the .psf file are (so far as I can tell) in exactly the same order
> as they are in the .pdb file.

you may want to override the atom "name" in the pdb file with
the atom "type" from the .psf file. to do this with VMD:

set mol [mol new "input.psf"]
mol addfile $mol "input.pdb"
set sel [atomselect $mol all]
$sel set name [$sel get type]
$sel writepdb "input-new.pdb"

hope that helps,
axel.

[Teodoro Laino]

Hi Chris,

a simple and additional comment to the one of Axel:
I'm sorry to say that just reporting an error without providing ALL
the files in order to reproduce the error is most of the times useless.
There are several reasons why you may have hit into this problem,
even a possible bug in the code.
Unfortunately if we don't work on the same files and cannot reproduce
the same error we can't say too much. I highly suggest to all the
people asking for help to provide ALL input files (i.e. input + all
the additional files necessary to run it).

cheers,
teo