Search Mailing List Archives

Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

Integration of terms into SQL db

Jacob Koehler jacob.koehler at
Sat Apr 20 08:37:32 PDT 2002


i developed a system where i had to solve similar problems:
please have a look at
(just click on login to enter).

> Chris Larsen schrieb:
> Hello GO freinds,
> We at Cognia are developing biological databases. After a long effort
> to make sure that the Java/SQL structure is up, stable in Windows NT,
> and can readily accept curated proteins, complexes, interactions,
> genes, and compounds live over the web from remote literature-reading
> curators, we have reached a point where the apparently difficult task
> of incorporating GO terms into our offerings has to be
> addressed. Previously I had tried to digest the terms from the huge
> list, by hand, and actually made good progress in producing 700
> ubiquitin-specific terms. BUT,
> This has its problems:
> 1    Does not address the terminology for the substrates of
> Ubiquitylation (80% of the cell?!)
> 2    List is static and does not evolve
> 3    List must be hard coded into a JSP page or as table values in SQL
> 4    Modifying the list is ridiculously cumbersome.
> Question:
> Does anyone have a strong method for incorporating terms?
> such that:
> 1    The evolution of GO is coupled to the database "live"
> 2    The huge number of terms is navigable to a curator when inputting
> data

i solved the two aforementioned problems storing the data in a
databases, not in the JSP. in oracle i used oracles proprietary "connect
by prior" statement for treeprocessing. some similar functionality could
be achieved by using SQL99s recursive SQL queries (which seems to work
in postgress now).

> 3    Terms changes can be back corrected, after the curator has used a
> term that becomes known as antiquated (aliasing?)

this was not on my list of requirements.
however, if you try to keep the data in the database, and process each
frontend page against the database, you could solve this by using an
apropriate database schema.


> We also might like to have, or know whether to generate if it hasnt
> been done, a live converter from EC to GO (to Cognia) numbers.
> (Biojava module?)
> If theres any help to be had here, please let me know. If we could
> automate this fully, that would be great, but our solutions right now
> are not up to large scale curation standards.  I suspect the right
> answer will be painfully simple and elegant, and in fact that's
> probably the one we want.
> Thanks,
> Chris
> ----------------------------------------------
> Christopher Larsen, Ph.D.
> Product Manager
> Cognia Corporation
> 117 E 55th St.
> New York, NY 10022
> e: clarsen at
> w: 212-331-7840
> h:  718-777-6558
> c:  347-837-0727
> f:   801-459-8850

     Jacob Koehler
     D5-117, AG Bioinformatik
     Universität Bielfeld
     PF 100131
     D-33501 Bielefeld
     tel: 0521-106-5296
     jacob.koehler at

This message is from the GOFriends moderated mailing list.  A list of public
announcements and discussion of the Gene Ontology (GO) project.
Problems with the list?           E-mail: owner-gofriends at
Subscribing   send   "subscribe"   to   gofriends-request at
Unsubscribing send   "unsubscribe"  to  gofriends-request at

More information about the go-friends mailing list