Re: [Psidev-ms-dev] PSI-MS WG ccall Tue 9am PDT, 12n EDT, 4pm GMT

[Eric D. <ede...@sy...>]

Hi everyone, here are my notes from the ccall, thank's for attending.

Regards,
Eric

------------------------------------------------------------------------
--------------------------

Open issues with dataXML

- Need documentation of each element and attribute
  Mike Coleman  2007-02-06
  Agreed. Does Kent have some start of documentation from
current/previous XMLSpy docs?

Eric will contact Kent directly

- Need more specifics on the desired numerical formats (e.g., IEEE,
etc.)
  Mike Coleman  2007-02-06
  To be discussed in Lyon

- Should there be "count" attributes?
  Frederik Levander  2007-02-06
  This is useful for some parsers. Probably leave in place. Discuss in
Lyon

- Should file index be within the file or in a separate file?
  Frederik Levander  2007-02-06
  The implementation of dataXML to be presented at Lyon will include
this information in the same file. Further discussion deferred to the
Lyon meeting The question was there: Should the file name / checksum
information be held in a separate file?

- Encoded filename should be element content instead of an attribute, so
that CDATA could be used
  Alex Masselot  2007-02-07
  To be discussed in Lyon

- Should controlled vocabulary term values also be element content
instead of attribute so that CDATA could be used?
  Alex Masselot  2007-02-07
  To be discussed in Lyon

- How will we handle multiple charges for parent peak, and fragmentation
peaks?
  Alex Masselot  2007-02-07
  This is already handled - mutliple cvParams of the same type can be
used to annotate one peak / spectrum.

- How can we allow two cv terms to be linked, such as concept with value
and the associated units (a separate cv term). (e.g. "collision
energy"=3D35.0 & "energy units"=3D"joules")
  Phil Jones  2007-02-27  (followup by Angel and Kent 02-28)
  Phil to post to list for further responses

Phil suggested 2 possibilities: <annotation> and additional attributes
Angel suggested doing the FuGE way. Maybe overkill? Difficult for
software to handle?
Do we really need having both or could we encode the units within the
terms, e.g. "collision energy in joules"
Most other relevant CVs don't seem to encode units as part of the term
Defer this to Lyon? Try to include the CV folks in a conference call
during meeting
Do we need to have specific types of relationships between linked terms,
like "has_units"?

- People still wrinkle their nose when hearing the "dataXML" name.  We
have a suggestion on the floor to rename to "mzDataXML".  Comments?

- Make sure dataXML web site is up to date as can be

- DONE Add link to XMLSpy-generated documentation at:
=20
http://gelbank.anl.gov/schema/documentations/dataXML0.11/dataXML0.11.htm
l

- Get the indexing wrapper schema working properly
  Jayson Falkner  2006-11-15

- Make sure that Karl Klauser is invited to be involved

Eric will send email.

- Do we support properly the spectrum "library" use case?
  Karl Klauser 2007-01-18
  dataXML is supposed to be for MS instrument raw data, not interpreted
data, i.e. with assignments. That is what analysisXML is intended for.

Still open questions here. If analysisXML doesn't includes spectra, this
would be tricky.

- Examine the mapping with MIAPE. Do we support everything MIAPE
requires?
  Pierre-Alain Binz  2007-03-30
  This will be addressed in Lyon

- Revisit the chromatogram use case and develop a good example


Controlled Vocabulary Issues:

Trish sent out a new version of the CV this morning.
New terms from Dave Horn added
Will start keeping and posting release notes
Pete should start working on this new version now
Pete, Trish, and I will iterate on CV a little and there will be another
ccall in 1 week exactly to discuss further
We should post the OBO file soon on OBO.

We will meet again in a week roughly, but Pete can't do next Tue 9am

- DONE. Add hyperlink to the current CV on the web site
  http://psidev.sourceforge.net/ms/xml/mzdata/psi-ms-cv-latest.obo

- Should we change the ontology namespace to PSI-MS?
  Trish Whetzel  2007-03-27
  Was suggested in Washington already and we decided no?

- What is the overlap/division between the PSI-MS CV, the main PSI CV,
OBI?
  Trish Whetzel, Luisa Montecchi  2007-03-30
  To be sorted out in a special working session in Lyon?

- Should we even have an InstrumentIdentifier (local to lab) term at
all?
  Trish Whetzel  2007-03-19

- What is required from vendors?
  Pierre-Alain  2007-04-03

- How do we deal with the common PSI CV?
  Pierre-Alain  2007-04-03

- What is the overlap with OBO ontology "ProPreo"?
  ProPreo: A comprehensive proteomics data and process provenance
ontology
  http://lsdis.cs.uga.edu/projects/glycomics/propreo/
  Eric Deutsch  2007-04-03

Trish says there was a quick review of this ~ 8 months ago
ProPreo is broader but not deep enough

- Try to reconcile the latest instance document and the current PSI
ontology:
  Trish Whetzel  2007-03-30
  Eric Deutsch responded 2007-04-03
  (full exchange not repro'ed here)
  Pete or someone trying to resolve based on discussion thus far and
identify unresolved?