Re: [Psidev-ms-dev] mzML 0.99.0 comments

[sneumann <sne...@ip...>]

On Mi, 2007-10-17 at 12:49 -0700, Brian Pratt wrote:
...
> something the current proposal largely flinches from doing.  As currently
> proposed, mzML feels like a big step backwards.

Hi,

greetings from one of the "lurkers" on this list.
We are operating a number of different MS.

Currently, we have used Eclipse EMF to auto-generate=20
Java classes from the mzData.xsd, and from there=20
we connect to a database, using an auto-generated schema
through an Object-relational Mapping ORM. The raw=20
data is read by the RAMP parser inside the Bioconductor XCMS package.

I have the feeling that a data model with very little structure
and a well-structured Ontology would put a lot of burden=20
on tool and database developers.=20

I expected mzML to be mainly a merger of mzXML and mzData,
keeping the best of both worlds, and improving vendor=20
and tools support for a merged standard. In that light=20
I followed the Index, Binary and Wrapper Schema discussion,
not responding because I saw that whatever way mzML settled,
I'd be able to adopt by ignoring those features or modifying=20
our tools. At the beginning of the mzML (when it was called dataXML)=20
discussion I also remembered the idea of having a place to store
the Chromatograms, I am not sure what happened to this.

Starting with the CV discussion I felt that mzML is drifting away
from its mz[Data|XML] parents. The rationale behind this discussion is=20
to keep up with ever-changing requirements.=20

But hey, mzData started in 2005, and will likely be applicable=20
to the majority of use cases another (at least?) 1-2 years.=20

I am not sure whether those use cases not covered by mzData=20
can easily be covered with mzML+complexCV, but for a speedy=20
adoption by both vendors please keep simplicity in mind.
Remember people will be writing mzML readers in Java, C++,
C# and Mono, perl, Bioconductor, Python, ... and It might turn=20
into a bad reputation for mzML if these implementations
are buggy and/or incomplete merely because mzML tries to
do too much and people end up hacking the parsers=20
just for their own machine and use case.

Yours,
Steffen

--=20
IPB Halle                    AG Massenspektrometrie & Bioinformatik
Dr. Steffen Neumann          http://www.IPB-Halle.DE
Weinberg 3                   http://msbi.bic-gh.de
06120 Halle                  Tel. +49 (0) 345 5582 - 1470
                                  +49 (0) 345 5582 - 0
sneumann(at)IPB-Halle.DE     Fax. +49 (0) 345 5582 - 1409