preamble.txt

   Open Bio Sequence Database Access.


 The Open Bio group of open source bioinformatics projects, being
BioPython, BioJava, BioRuby and BioPerl have agreed on a
cross-language and cross-platform specification for accessing
biological sequence databases. This specification has been
incorporated into the four projects and provides bioinformatics
application developers seemless access to databases. The specification
allows:

  (a) a global "fallback" access to databases, to allow applications 
installed with no bioinformatics support on site to retrieve sequences

  (b) the ability to configure locally (to a system) or personally (to a user)
access to databases, allowing customisations of the database set up and 
on-site specific databases to be accessed

  (c) A number of implementations of datastores/data fetching protocols which 
work portably between all three main projects (Bioperl, Biojava and Biopython)
including flat file indexes, simple web services, more complex, SOAP based
databases and relational schema databases. These implementations are accurately
specified and each project has provided read access to these implementations
and at least one project has write or build systems for them.

  All implementations are based on working systems in at least one of
the projects and are expected to be stable and robust.
 

  The rest of this document is broken into parts being

  (a) the definition of the registry, being the configuration of where the 
database services exist

  (b) the definition of each of the implementations being

    - flat file indexing (two implementations are defined)

    - simple web fetching (biofetch)

    - CORBA, BSANE compliant access

    - SOAP based access, based on XEMBL

    - and relational database, based on biosql access


It is hoped that other projects can adopt these access methods and therefore
system administrators and bioinformatics database maintainers on site can
simplify the management of databases.