I have taken TIGR's Apollo chado database adapter, developed by Johnathan
Crabtree, and adapted it to work with FlyBase chado. It is currently
read-only. Writing in FlyBase is done through GAME XML to chado XML to
chado. Directly writing back to the database is probably a ways off, as
it will take a bit more effort and there are other priorities. Also policy
issues need to be hammered out with direct writeback (e.g., permissions,
transactions, backup, etc.). If this is something you need sooner, please
let us know.
The jdbc adapter is included in the latest release of apollo. A user can
query FlyBase with gene name or a scaffold id. If you would like greater
querying capability, please let me know.
At this point the adapter is not very configurable from the configuration
files. There is a single configuration flag (in the style file) that flips
it between TIGR and FlyBase. TIGR is using an old version of chado
(pre-Sequence Ontology), while FlyBase is using a more recent version, so
the queries between the two can vary a bit.
There is a chado-adapter.cfg file (in the apollo/conf/ directory) where
one can plug in one's own chado database. In this file, the jdbc url,
database name, and login (optional) can be specified.
If you are using a chado database that is identical to FlyBase (Sequence
Ontology compliant) it should in theory work for you. Given the flexible
nature of chado, I wouldn't be surpised if there are differences.
Minimally, you will probably have different result sets to retrieve--that
should probably be one of the first things to configure. There are
innumerable different places where the adapter could be configured. I'm
taking a lazy, on-demand approach to configuring, rather than trying to
surmise where other groups chado's will diverge. In other words, give the
adapter a try on your chado. Let me know where it went awry and where
you need the adapter to be configured, and I will try to accomodate.
Looking forward to your feedback,