Merged triples creating "orphaned" axioms/evidence about gocamgen HOT 6 CLOSED

Well, I do notice that there are some axioms in this OWL that already use my proposed solution:

[ rdf:type owl:Axiom ;
   owl:annotatedSource <http://model.geneontology.org/66ab4fd6-68b2-4e3c-aa19-88199eca825e/91d1e1f5-8d54-4f09-aa47-f2e3dd12c672> ;
   owl:annotatedProperty <http://purl.obolibrary.org/obo/RO_0002333> ;
   owl:annotatedTarget <http://model.geneontology.org/66ab4fd6-68b2-4e3c-aa19-88199eca825e/1e478989-604a-4cfb-ac5f-78324e0ad69d> ;
   <http://geneontology.org/lego/evidence> <http://model.geneontology.org/66ab4fd6-68b2-4e3c-aa19-88199eca825e/27f7ae19-d188-4f82-ad55-3994933c147c> ,
                                           <http://model.geneontology.org/66ab4fd6-68b2-4e3c-aa19-88199eca825e/89c7879d-9f93-4a1c-9f90-318b2e30c05e> ;
   <http://purl.org/dc/elements/1.1/contributor> "http://orcid.org/0000-0002-6659-0416" ;
   <http://purl.org/dc/elements/1.1/date> "2012-10-04" ,
                                          "2012-10-05" ;
   rdfs:comment "MGI MGI:87859 enables GO:0005515 MGI:MGI:1099567|PMID:8861907 ECO:0000353 UniProtKB:Q03173  20121005 MGI" ,
                "MGI MGI:87859 enables GO:0005515 MGI:MGI:1349267|PMID:10567556 ECO:0000353 UniProtKB:Q9QZK7  20121004 MGI"
 ] .

For MGI:MGI:87859 to GO:0005515 annotations.

So something else must be going on to explode out the evidence for the MGI:MGI:87859-to-GO:0004713 associations into multiple axioms.

from gocamgen.

@dustine32 I've experienced this a lot in my development. Somehow you ended up with an evidence instance that is not connected to anything else. When I have done this to myself, it has usually been related to trying and failing to delete a relationship that I have added to the model. Delete is a pain as you need to cascade it through all the evidence nodes etc. If you send me a full model that contains the problem maybe I can help diagnose.

from gocamgen.

Thanks @goodb ! Here's the model for MGI:MGI:87859. I can email you the original TTL file if you're interested.

As far as I know, my code shouldn't be deleting relations when building the graph that gets submitted to rdflib (builds the TTL). I try to first search and reuse axioms and create if they don't exist. But I think my add_evidence code is redundantly emitting the axiom for some of these evidences. I probably need to break some of the emitting responsibility out of this add_evidence function. Though I haven't yet followed this through the debugger.

from gocamgen.

I notice there are bunch of assertions with no evidence. Perhaps evidence nodes are getting created but not linked up? e.g.

gomodel:66ab4fd6-68b2-4e3c-aa19-88199eca825e/9e18e9f5-67b0-4fd0-8165-be66607f7acc / RO:0002333 / gomodel:66ab4fd6-68b2-4e3c-aa19-88199eca825e/e46feaf1-2984-43d0-a3ed-d292cead6ef0

from gocamgen.

Aha! Thanks @goodb this triple pointed me to the likely weirdness going on with my extension code. I just switched from reusing the extension logic from my PomBase models to something much simpler and this problem appears to have gone away.

from gocamgen.

Should be fixed by e6c0589

from gocamgen.