[Psidev-pi-dev] PSI Fall 2006 meeting summary

[Angel P. <an...@ma...>]

The PSI Fall 2006 working group meeting in Washington, D.C. was a 
rousing success story for the Mass Spec and Proteomics Informatics 
working groups. First and foremost, the working group chairs would like 
to thanks everyone in attendance, as all put forth an unprecedented  
effort across the numerous activities.

For those of you that were not able to attend, here are some highlights:

The general philosophy was "get things done" and to that end, the 
attendees were split up into multiple smaller groups with specific 
deliverables. These included a thorough review and sign off of the mass 
spec engine spreadsheet, work on the much talked about merge of the 
mzData and mzXML formats, the analysisXML UML model, and the beginnings 
of the ontology for use with analysisXML.

Fantastic progress was made on all of these fronts:

mzData/mzXML
-----------------------------------
The document of the difference between the formats produced and 
presented by Kent last meeting was used to begin merging of the schema. 
The data arrays where worked out for the most part, as was annotation 
and generalization of the instrument, protocol and parameter 
annotations. The work is slated to be finished by the end of the year, 
mostly by use of email and a satellite meeting in Seattle. Join the next 
conference call for further details. Use cases of instrument modes (MRM, 
LC-MS, LC-MALDI MS/MS, neutral loss scans, etc) were also submitted to 
be mapped on the model.

The ability to computationally validate MS interchange data for MIAPE 
compliance was also discussed.  The working group is considering 
encoding MIAPE concepts and terms into a CV or ontology, which could be 
used for
future software validation of XML instance documents.

AnalysisXML UML model
-----------------------------------
Modeling was started using FuGE as a basis. A provisional model was 
created and turned into XML schema using the AndroMDA tools. Since time 
was limiting factor, the development effort did not pay careful 
attention to documentation and diagram formatting, thus the model is 
undergoing a bit of clean up before release to the rest of the WG.

AnaysisXML content and CV
-----------------------------------
Based on the search engine spreadsheet generated during this summer, the 
group did a rigorous review of the content that AnalysisXML should carry 
and mapped the current search engine outputs with MIAPE requirements and 
MCP guidelines for reporting mass spec search engine parameters, even 
adding a few that where missing. A first proposal for CV terms has been 
generated. The group, was able to agree on a large subset of the 
parameter names and meanings, which are being added to the PSI ontology. 
Vendors will be asked for vendor-specific terms. Not covered was 
quantitative parameters. Currently Jim Shofstal is cleaning up the 
document prior to sending to the rest of the WG.

A discussion was raised about a possibility to homogenize the Accession 
codes used in the various engines. A main difficulty comes from the 
interpretation of the fasta header lines by the various tools. A 
proposal was to study the fasta format generated by Phenyx that is 
structured in a way it clearly labels different information types such 
as AC, Description, taxonomy, PTMs, etc. Stay tuned for details 
regarding this.




Once again, thanks everyone for being in attendance and for all of the 
hard effort allowing such an amazing amount of progress in the short 
time we had together.

Cheers, from the PSI-MS and PSI-PI working group chairs and secretary:
Pierre-Alain Binz
David Creasy
Phil Jones
Randy Julian
Angel Pizarro