[Biojava-l] nested features

Martina boehme at mpiib-berlin.mpg.de
Thu May 19 08:18:09 EDT 2005


Thanks for your interest, Richard,

I should have made this clearer: yes, I'm expecting the second output:

retrieving AF100928
AF100928 contains 1 features
Source: ncbi Type: gen contains: 2
Second Level: Source: AIF Type: siRNA contains: 1 Second Level: 
Source: AIF 5`450 / AIF 3`682 Type: Amplicon contains: 0

I am trying to upgrade to MySQL 4.1.12 (was 4.1 alpha before (maybe it 
was a problem there)) and it is not working right now, but I will post 
the results of the sql statements as soon as the db is up again.

Martina

Richard HOLLAND wrote:
> Just to get this straight, if I follow your code correctly, then the second example output you should be seeing is the correct one. Could you confirm that this is what you are expecting? At least then it gives us something to work towards as a definite known correct answer.
> 
> It would also be helpful if you could provide, for each of the two cases, the contents of the seqfeature and seqfeature_relationship tables. You don't need to provide them all. Just find out the bioentry_id of the inserted sequence (look it up in the 'bioentry' table using the 'name' column to look for your sequence name), then record all the rows from seqfeature with bioentry_id being that value, and all rows from seqfeature_relationship where either subject_seqfeature_id or object_seqfeature_id matches any of the seqfeature_id values you just got from the seqfeature table.
> 
> In other words (assuming Oracle, but similar queries will work elsewhere):
> 
> 	select seqfeature.*
> 	from	seqfeature, bioentry
> 	where	seqfeature.bioentry_id = bioentry.bioentry_id 
> 	and	bioentry.name = 'AF100928';
> 
> 	select seqfeature_relationship.*
> 	from seqfeature_relationship, seqfeature, bioentry
> 	where	(subject_seqfeature_id = seqfeature.seqfeature_id 
> 		or object_seqfeature_id = seqfeature.seqfeature_id)
> 	and	seqfeature.bioentry_id = bioentry.bioentry_id 
> 	and	bioentry.name = 'AF100928';
> 
> Also could you change all the "Iterator<Feature>" references in your test code to just "Iterator" as BioJava does not use Java 1.5 internally and I'm not sure what effect genericising the iterator might have (Java 1.5 can do funny things due to the effect of what they call erasure). This probably won't fix it, but it's worth a try.


More information about the Biojava-l mailing list