BioSQL is a Bioperl-based system for storing record-oriented biological objects, including GenBank/EMBL sequences and PubMed records, in a relational database. GBrowse provides support for viewing sequence annotation data stored in BioSQL. The support is functional, but not heavily tested, so it may still contain bugs.
Make sure that Gbrowse is correctly installed.
Make sure that you have the latest version of Bioperl-db.
Make sure that you have the latest version of Bioperl-schema.
You need to have a database (MySQL, Postgresql, Oracle) that is supported by the Perl DBI interface.
If you plan to use MySQL, run create_mysql_db.pl (it is included in the bioperl-schema distribution). Other databases would use similar SQL commands.
Take a Genbank file and call the script load_seqdatabase.pl from the bioperl distribution to load the data. For example,
load_seqdatabase.pl --host somewhere.edu --dbname biosql \
--namespace bioperl --format genbank \
your_genbank_file
The value of ``namespace'' is arbitrary.
Point a gbrowse configuration file to this database. See include 06.biosql.conf for an example. The only nontrivial parameters in it are ``namespace'' and ``version''. For namespace, use the same value that you specified when you uploaded data. The version number is contained in the Genbank file.
Until these files are uploaded into the bioperl CVS, put the attached files on your system in a directory accessible to Perl, for example, into /usr/lib/perl5/site_perl/5.8.1/
Please send requests for help to simonf@cshl.edu or <a href=``mailto:gmod-gbrowse@lists.sourceforge.net''>gmod-gbrowse@lists.sourceforge.net.
Vsevolod (Simon) Ilyushchenko simonf@cshl.edu