Re: [h5md-user] Tuples and topology, the return

[Pierre de Buyl]

On Thu, Jul 10, 2014 at 03:39:04PM +0200, Konrad Hinsen wrote:
> Felix Höfling writes:
> 
>  > What I understand is that PDB is still the prevailing format for protein  
>  > data, but nobody likes it. Does the PDB offer alternatives yet?
> 
> The PDB has been promoting mmCIF as a replacement for the PDB format
> for about 15 years. They also jumped on the XML bandwagon and
> introduced PDBML as an alternative. Both mmCIF and PDBML files contain
> the full information stored in the database, much more than the
> PDB files can handle. And both are very well documented.
> 
> mmCIF has had a big acceptance problem, because nobody wanted to
> change their software. But the problems with the PDB format are very
> real as well. The PDB has finally decided to drop support for PDB
> files in 2016.
> 
>  > I wasn't aware that the PDB files come with a bundle of
>  > complementary file formats. This information, of course, should be
>  > combined in a single H5MD file. If other short-comings of PDB
>  > could be fixed on the way this would be great. I think it is a
>  > worthwile goal for H5MD to potentially become a standard container
>  > for protein data.
> 
> An mmCIF file contains everything. And H5MD cannot handle any of the
> information in a PDB entry. There is no overlap at the current stage
> of H5MD.

There is. Storing properly [1] the position data is already a nice feature of
H5MD. In addition the positions, we are also able to store successive sets of
positions (an abuse that is sometimes made for PDB), the species of the
particles.

There is clearly a gap (some would say a 'quantum jump') between H5MD and PDB
but given that H5MD 1.0 is out I expect (or hope) that the possibilities to
reduce that gap will appear out of the practice.

P

[1] I write "properly" just to flatter "The H5MD Developers" :-)

> 
> Konrad
> -- 
> ---------------------------------------------------------------------
> Konrad Hinsen
> Centre de Biophysique Moléculaire, CNRS Orléans
> Synchrotron Soleil - Division Expériences
> Saint Aubin - BP 48
> 91192 Gif sur Yvette Cedex, France
> Tel. +33-1 69 35 97 15
> E-Mail: research AT khinsen DOT fastmail DOT net
> http://dirac.cnrs-orleans.fr/~hinsen/
> ORCID: http://orcid.org/0000-0003-0330-9428
> Twitter: @khinsen
> ---------------------------------------------------------------------
> 

-- 
-----------------------------------------------------------
Pierre de Buyl
KU Leuven - Polymer Chemistry and Materials
T +32 16 327355
W http://pdebuyl.be/
-----------------------------------------------------------