Proteomics Standards Initiative

1. What is the potential use of a public archive of mass spectrometry data?
   1. Confirmation of hypothetical proteins in sequence databases
   2. Making raw data available for third-party analysis and verification of experimental results. Example: Links from identified 2D gel spot to original spectrum.
   3. Re-analysing existing spectra with new approaches, or on the basis of new data, for example updated versions of sequence databases.
   4. Improvement of future identifications by spectra comparison for defined samples?
   5. Provision of reference sets from defined samples?
   Aim: Describe intended uses and define their requirements, possibly split into consecutive development phases.
2. What data should be stored?
   1. raw, instrument-dependent data
   2. instrument-independent data after noise reduction ("peak lists")
   3. interpreted data: Identifications (in which form)
   4. Scope of auxiliary data, for example source system: Should the auxiliary data allow to recreate the experiment, or just compare data from different experiments.
   5. Elements of auxiliary data:
      1. source system: organism, cellstage, ..
      2. machine type
      3. experimental setup
   6. data context: individual identifications of proteins versus "proteome of cell X"
3. Which existing systems/controlled vocabularies/ontologies can be used, where do we need additional systems?
4. Representation of proposed data format: XML/UML?
5. Beyond data format: Data collection
   1. Database setup:
      1. central database,
      2. federated database with data exchange,
      3. federated database with common query interface?
   2. data exchange:
   3. flat/XML files
   4. Define availability levels:
      1. free
      2. free to end user (no redistribution)
      3. only "detail" access via tools
      4. private
      5. Data ownership, update policy.
   5. Quality control:
      1. accept only published data?
6. Synchronised effort to create reference sets?