Extra unwanted element from a query on PC2v5

[GoogleCodeExporter]

Use the web service to get the BiochemicalReaction:

http://purl.org/pc2/5/BiochemicalReaction_4388d4a04f2b3f5f6085c601293ea5b1

In the returned model, you will notice that there is also a Catalysis object:

http://purl.org/pc2/5/Catalysis_9eac00623407371a693efb800ef6aad6

which controls this BiochemicalReaction. The query result should not contain 
this Catalysis, as we don't "complete" conversions with their controls (the 
reverse is true).

Original issue reported on code.google.com by ozgunba...@gmail.com on 14 Oct 2014 at 4:54

[GoogleCodeExporter]

This is related to issue #173 (a special fix, originally for PANTHER pathways), 
where we, in addition to using the Completer, also add missing Controls to the 
model returned by a "/get" (fetch by URI/ID) query.

I (we?) thought that was a good idea. 

Again, note, this is only in effect for our "/get" queries.

Thinking...

Original comment by rod...@gmail.com on 16 Oct 2014 at 6:42

• Changed state: Accepted
• Added labels: Type-Defect, Priority-High, Product-cPath2

[GoogleCodeExporter]

I think we should move this fix to the panther normalizer. As far as I can see, 
the fix in /get query does not apply only to panther data. This fix for panther 
may break other databases if they put a conversion into a pathway but not some 
of its activator(s), on purpose.

If we decide to keep this fix (I hope we won't), then there is another issue 
with the fix. The query of the biochemical reaction that I mentioned in the 
issue returns the Catalysis object, but it does not bring its controller. So it 
looks like this enriched result should also be "Completed".

Original comment by ozgunba...@gmail.com on 16 Oct 2014 at 8:01

[GoogleCodeExporter]

Aye, fair enough. I am removing this code, i.e, the fix used for issue #173 
(in fact, PantherCleanerImpl has been already updated to address the same 
issue). 

Original comment by rod...@gmail.com on 16 Oct 2014 at 8:38

[GoogleCodeExporter]

This one is now fixed (in the test PC2 v5 instance).

Original comment by rod...@gmail.com on 20 Oct 2014 at 9:24

• Changed state: Fixed