[opdev] various questions

2 views
Skip to first unread message

Pierre Lindenbaum

unread,
Nov 12, 2008, 9:24:18 AM11/12/08
to foglio, opero...@googlegroups.com
Hi Mario

Various questions:
1) on cabernet : is '/usr/local/BerkeleyDB.4.6' the current library
used by the database in /operon/
2) Can I recompile it with the java option ?
3) Can I find all the databases described in your document
Operon_DB_Schema on cabernet (I guess no)
4) For each instance of each database, could you please give me the
relative path: for example "dbs/gen/xalias.bdb" for 'Xalias"
5) still want me to go at Evry for a tutorial ? :-)

On my side, I've worked on the my code to specialize the databases
(e.g. specific methods if they allow or not duplicate keys )

Pierre

Mario Foglio

unread,
Nov 12, 2008, 11:47:16 PM11/12/08
to opero...@googlegroups.com
Hi Pierre,

What we are going to do in cabernet is to create the new operon dbs as in
the latest schema we have, but we need to keep 4.6 running while we
build 4.7 with java. With the new version, the idea is to replace the cgi
work I did in 4.6 with your java approach. As a minimum, what Mark
expects (in 4.7) is what we see now in 4.6 (regulon.cng.fr) since there
are some key people already using it. It does not need to be exactly
the same, but the same functions and layouts are expected. In addition,
the new version will start queries via the controlled vocabularies and not
via xalias. We will need to discuss this in order to come out with a good
use of it. To start I will send you an example of how this can be used.
However, I am waiting for the new release of the UMLS sources that
it is coming out this month. Last update was in April, and this update is
important because it will include the Foundational Model of Anatomy,
and we need that vocabulary.

For the relative paths in the new version, you can find them in the schema
I send you. The paths will have these two forms:
/operon/bdb/dbs/<domain_name>/<db_name>
/operon/bdb/dbs/<domain_name>/<species_initials>/<build_version>/<src_name>
To find out which ones are species specific paths you have a 1 in the
first column of the schema (column named PL).
The data domains are projects, DNA, genes, proteins, and hypergraphs.
The previous "systems" name is renamed with hypergraphs.

For your tutorial that will be great! We will need to set-up a date. I am
going to ask Simon to find out who may be interested to be present.
Today (and yesterday also) I am off work. I have to take care of kids
stuff. I will look my emails again this afternoon.

Mario

Pierre Lindenbaum

unread,
Nov 13, 2008, 10:01:08 AM11/13/08
to opero...@googlegroups.com, foglio
> With the new version, the idea is to replace the cgi
> work I did in 4.6 with your java approach.

ok, that's a clear milestone for me.
e.g. in attachement, here iI implemented the 'SEQMAP' list in my
browser today. It needs a few CSS tricks but it works fine.

> To start I will send you an example of how this can be used.

ok, that would be cool


> For your tutorial that will be great! We will need to set-up a date.

ok, you can check in my calendar.

Warning, there is a strike next week
http://fr.news.yahoo.com/2/20081113/tfr-sncf-deux-syndicats-appellent-la-gre-b7b6525.html

A+
Pierre

Screenshot.png
Reply all
Reply to author
Forward
0 new messages