[gmx-developers] Free Format gro File

[David van der Spoel]

On 2012-10-02 09:13, David van der Spoel wrote:
> On 2012-10-02 00:59, Jicun Li wrote:
>> You are right, Mark. Thank you very much for your suggestions.
>>
>
> I think it is better to adopt another format altogether that is
> supported out of the box by other programs as well.
>
> SDF is such a format:
>
> http://en.wikipedia.org/wiki/Chemical_table_file
>
> It would be great to have native SDF support in gromacs.
>

OK, just to comment my earlier message: apparently SDF is also a fixed 
format file, at least in the babel implementation (example below). It is 
in Angstrom with 4 digits, and hence 100 times more resolution than gro 
files (10 times more than pdb). The maximum coordinates are 99999 
Angstrom or 9999 nm or < 10 micron. This is probably OK for a while for 
atomistic simulations but not in the long run for coarse-grained models.

However, another drawback is that babel/the format only accepts elements 
rather than atom names.

Sample:

cyclohexa-13-diene.pdb
  OpenBabel10021209253D

  14 14  0  0  0  0  0  0  0  0999 V2000
    -1.1900   -0.7290    0.2420 C   0  0  0  0  0  0  0  0  0  0  0  0
    -1.1910    0.7270   -0.2420 C   0  0  0  0  0  0  0  0  0  0  0  0
     0.1140   -1.4190   -0.0630 C   0  0  0  0  0  0  0  0  0  0  0  0



>
>> Jicun
>>
>> 2012/10/1 Mark Abraham <Mark.Abraham at anu.edu.au
>> <mailto:Mark.Abraham at anu.edu.au>>
>>
>>     On 2/10/2012 8:22 AM, Jicun Li wrote:
>>
>>         Maybe the fixed-format gro file is easier to code with in native
>>         F77, but it's difficult to prepare with other molecule building
>>         software.
>>
>>
>>     Sure, but once you've picked a file format, you're stuck with it for
>>     a long time.
>>
>>
>>         I would like Gromacs to read in the structure exactly, at least
>>         as exact as g96 file. But I do not like to use g96 file because
>>         there is no atom name in it. I can not use other software to
>>         view the structure.
>>
>>
>>     There's a separate library of code available on the website for
>>     linking to external projects that want to write GROMACS-readable
>>     trajectory file formats. That might be useful for you.
>>
>>     Wanting to write the atom names in the trajectory file is wasteful
>>     if you have a big trajectory. For this reason, tools like GROMACS
>>     and VMD can read a structure file and then import a trajectory file
>>     that lacks atom information but whose atom ordering matches.
>>
>>     You'll be much better off using tools that suit the job at hand:
>>     * When you want to view the structure, write a structure file
>>     because it has atom names and stuff.
>>     * When you want a high-precision trajectory, write a trajectory file
>>     because it has high precision.
>>     * When you want to view a trajectory, import the latter onto the
>>     former. You may wish to reconsider your choice of viewer if you
>>     can't do this :-)
>>
>>
>>     Mark
>>     --
>>     gmx-developers mailing list
>>     gmx-developers at gromacs.org <mailto:gmx-developers at gromacs.org>
>>     http://lists.gromacs.org/__mailman/listinfo/gmx-__developers
>>     <http://lists.gromacs.org/mailman/listinfo/gmx-developers>
>>     Please don't post (un)subscribe requests to the list. Use the www
>>     interface or send it to gmx-developers-request at __gromacs.org
>>     <mailto:gmx-developers-request at gromacs.org>.
>>
>>
>>
>>
>
>


-- 
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205.
spoel at xray.bmc.uu.se    http://folding.bmc.uu.se