[Biojava-l] Re: [BioSQL-l] _removeSequence
hollandr at gis.a-star.edu.sg
Sun Jun 26 11:11:30 EDT 2005
The revamped BioJava/BioSQL classes will expose the rank to the user for all tables which have ranks.
From: biosql-l-bounces at portal.open-bio.org on behalf of Simon Foote
Sent: Wed 6/22/2005 12:15 AM
Cc: Hilmar Lapp; biosql-l-bounces at portal.open-bio.org; BioJava; biosql-l at open-bio.org
Subject: Re: [Biojava-l] Re: [BioSQL-l] _removeSequence
In fact you can, as rank is the field that allows this to happen. In
Biojava, currently it's just a linearily incremented number such that
you can have the same type and source IDs for a given bioentry.
For example, adding a Genbank entry with 10 CDS features for 1 bioentry
will give you identical keys for bioentry_id, type_term_id and
source_term_id, but will have a rank of 1 - 10 for each.
> That means, that I can't have 2 features refering to the same bioentry
> with the same type (= type_term_id)and source (=source_term_id) but
> different parent features because of the composite key bioentry_id in
> the seqfeature table? Or what does "rank" in that table mean (its part
> of that key), how can I get different ranks?
> Hilmar Lapp wrote:
>> The Biojava people will respond to this. Note though that
>> Term_Relationship is for storing subject-predicate-object triples of
>> terms, so I'm not sure why you want to use it for storing/associating
>> annotation. Maybe you meant bioentry_qualifier_value?
>> On Jun 21, 2005, at 3:10 AM, Martina wrote:
>>>> Yes. When you insert a sequence you must be prepared that when
>>>> inserting its ontology term or tag/value annotation the term may
>>>> already be present because another bioentry uses it too.
>>> Ok, the proper way is to catch the SQLException in BIOSQLFeature,
>>> test if it is a Dublicate key entry, get the identifier of the term
>>> (would that be the BioSQLfeatureId ?) and insert it in the
>>> term_relationship table? And there is no nice BioJava method for
>>> this, I have to do it "manually", like conn.prepareStatement(..) and
>>> stuff? BioJava spoiled me so!
> Biojava-l mailing list - Biojava-l at biojava.org
Institute for Biological Sciences
National Research Council of Canada
[T] 613-990-0561 [F] 613-952-9092
simon.foote at nrc-cnrc.gc.ca
BioSQL-l mailing list
BioSQL-l at open-bio.org
More information about the Biojava-l