question on biosql databases

[Dazhi Jiao]

Hi,

I am a student working on the phylr web service this summer. Right now
I am trying to setup a BioSQL database to test with. I wonder if there
are any kind of database available that could be the backend of
phylows. If so, is it possible for me to have a dump from the database
so I could import it into my database for testing?

Otherwise, if you could point me to data that are suitable for putting
in such a database, it'll also be very helpful.

Thank you!

Dazhi Jiao

[Hilmar Lapp]

Hi Dazhi,

there is a script called load_ncbi_taxonomy.pl that comes with BioSQL
and will load the NCBI taxonomy into it (though into a part of BioSQL
that isn't PhyloDB - but it can be moved to there using SQL). There is
also a script called load_itis_taxonomy that loads ITIS into the
phyloDB tables.

Finally, Bill Piel had some dumps of TreeBASE in a simplified schema
at the DB Interop hackathon. If you load that you can probably easily
move it into phyloDB tables using SQL commands. I'm copying Bill as to
where those dumps are available.

-hilmar

--
===========================================================
: Hilmar Lapp -:- Durham, NC -:- hlapp at duke dot edu :
===========================================================

[Dazhi Jiao]

Hi, Hilmar,

I just found that there is a PhyloDB extension of the BioSQL. I guess
that's what you referred to in your email. I am trying to map the
TreeBase data to the tables in that extensions, and some other BioSQL
tables. I found that the Study table in the TreeBase database has much
more granular information than the Reference table in BioSQL, and the
PhyloDB extension doesn't seem to have a table for such information.
So I guess I'll have to drop some data when doing the mapping. Please
let me know if there is anything I missed.

Do you or anyone know if the BioJava API has implementations for the
PhyloDB extension?

Thank you!

Dazhi

[Hilmar Lapp]

Hi Dazhi,

you will have to map the study-level information that you want to
preserve to metadata of the tree (i.e., entries in the
tree_qualifier_value and term tables). You can preserve as much as you
want, you just need to "invent" a vocabulary for it. You might start
with the one that Rutger put up for querying.

-hilmar