[gmx-users] support mmcif format?

[Albert]

probably there is some misunderstanding.

I didn't mean to store Gromacs trajectory to mmCIF format, but 
extracting a specific snapshot from Gromacs trajectory into mmCIF file.

It is no doubt that the .xtc or .tng format is the best format for 
Gromacs trajectory. However, when we use command:

trjconv -f md.xtc -s md.tpr -o test.pdb -sep -b xx -e xx

It will lead to many problems from the .PDB file if our system is too 
big.... By contrast if it is mmCIF or PDBX format, everything would be 
perfect.

Albert




On 11/23/2015 12:08 PM, Hannes Loeffler wrote:
> It's hard to see for me why someone would even want to use mmCIF as
> a trajectory format.  The format serves, primarily, the crystallography
> community to store a single set of coordinates together with a huge
> amount of meta data, most of which are probably rather irrelevant for
> trajectories.   (For archiving purposes this is of course interesting
> but no need to write this into every single trajectory, just like the
> topology is kept separate).
>
> I think for the simulation community the mmCIF format should be what the
> PDB format should always have been: a read-only file format to be
> converted into a format developers think is best for their particular
> piece of software.  Any future setup program should be able to read
> mmCIF but several libraries are already available to do exactly that.
>
> Cheers,
> Hannes.