Service provider Workgroup ¶

Here, we gather the major web service providers including BioMOBY projects and Japanese providers such as DDBJ, KEGG, PDBj, CBRC and other foreign providers like EBI and NCBI.

• Discuss current problems existing in each services:
  • Language dependencies
  • Naming conventions
  • Returned data type
  • Returned data size
  • Job managements
  • WSDL syntax violations
  • Timeouts
  • Sync and async
  • MIME attachments
  • Not implemented yet

• How to integrate non-MOBY web service providers?
  • What kind of effort is needed to compromise with BioMOBY framework?

• How to improve interoperability among the services?
  • Do you need proxy server to translate objects among services?
  • Object identity matching problem (LSID, BioMOBY)

• How to treat the needs for the REST services?
  • PDBj's REST API (prototype):  http://doc.pdbj.org/help.cgi?REST%20API
  • DDBJ's REST API:  http://www.xml.nig.ac.jp/tutorial/rest/index.html
  • EBI's REST API  http://www.ebi.ac.uk/Tools/webservices/tutorials/rest

Attending Service Providers ¶

• STRING/STITCH,  http://string.embl.de /  http://stitch.embl.de (Michael Kuhn)
  • scope of database: protein-protein and protein-chemical interaction networks
  • existing web-services: rudimentary/broken SOAP interface to retrieve network via PSI-MI
  • plans for API: query for the network (return PSI-MI 2.5), query for protein by identifier, synonym or sequence, retrieve orthologs/homologs, retrieve synonyms for protein, retrieve abstracts mentioning a protein
  • open questions:
    • Which ontologies should be used for the input (i.e., how to represent a protein or chemical so that other web services can deal with it)?

• DATABASE,  http:// ()
  • scope of database:
  • existing web-services:
  • plans for API:
  • open questions: