Failing QC checks (labels containing IRIs)

[ptgolden]

There are currently 6 violations of the sparql query label-with-iri-violation.sparql:

robot --catalog catalog-v001.xml verify -i tmp/merged-zp-edit.ofn --queries ../sparql/owldef-self-reference-violation.sparql ../sparql/iri-range-violation.sparql ../sparql/label-with-iri-violation.sparql ../sparql/multiple-replaced_by-violation.sparql ../sparql/dc-properties-violation.sparql -O reports
PASS Rule ../sparql/owldef-self-reference-violation.sparql: 0 violation(s)
PASS Rule ../sparql/iri-range-violation.sparql: 0 violation(s)
FAIL Rule ../sparql/label-with-iri-violation.sparql: 6 violation(s)
term,value
http://purl.obolibrary.org/obo/ZP_0137765,"http://purl.obolibrary.org/obo/CARO_0000010 hepatocyte amorphous, abnormal"
http://purl.obolibrary.org/obo/ZP_0149667,"http://purl.obolibrary.org/obo/CHEBI_15368 whole organism increased amount, abnormal"
http://purl.obolibrary.org/obo/ZP_0016006,"http://purl.obolibrary.org/obo/CARO_0000010 somite morphology, abnormal"
http://purl.obolibrary.org/obo/ZP_0023004,"http://purl.obolibrary.org/obo/CARO_0000010 myotome spatial pattern, abnormal"
http://purl.obolibrary.org/obo/ZP_0149670,"http://purl.obolibrary.org/obo/CHEBI_15368 liver increased amount, abnormal"
http://purl.obolibrary.org/obo/ZP_0141352,"http://purl.obolibrary.org/obo/CARO_0000010 notochord straight, abnormal"
PASS Rule ../sparql/multiple-replaced_by-violation.sparql: 0 violation(s)
PASS Rule ../sparql/dc-properties-violation.sparql: 0 violation(s)

Two of these are from a missing CHEBI term, which is solved by refreshing the imports. The remaining four are related to #63.

[ptgolden]

CARO is removed from the merged imports here:

zebrafish-phenotype-ontology/src/ontology/Makefile

Line 409 in 8f52fb8

remove --select "<http://purl.obolibrary.org/obo/CARO_*>" \

Should we add it back in until #63 is resolved? Or alternatively, we could manually add CARO:0000010 in zp-edit.owl.

[matentzn]

Can you trace back how CARO:0000010 entered the pipeline? It should not be used in any pattern at all, since it is deprecated. Did ZFIN bring it in?

[ptgolden]

Yes, it's from ZFIN. Four phenotypes use it:

https://zfin.org/ZDB-FIG-191230-1574#phenotype-data
https://zfin.org/action/phenotype/statement/644783648

https://zfin.org/ZDB-FIG-140523-38#phenotype-data
https://zfin.org/action/phenotype/statement/644959845

https://zfin.org/ZDB-FIG-170417-7#phenotype-data
https://zfin.org/action/phenotype/statement/644759795

https://zfin.org/ZDB-FIG-190723-877#phenotype-data
https://zfin.org/action/phenotype/statement/644851052

[ybradford]

@matentzn I'm looking at this to correct these annotations- I do not see anatomical boundary obsoleted in BSPO - that is where we are using this term from http://purl.obolibrary.org/obo/CARO_0000010. Can you please point me to where it is obsolete term in BSPO?

[matentzn]

Oh sorry, I misspoke. I did not mean to say CARO:0000010 is deprecated. I meant to say "CARO (the whole ontology) is deprecated". Unfortunately that has never been done officially, but it has been removed from Uberon: obophenotype/uberon#2614

The fact that the class still exists in BSPO is annoying, but unfortunately I cannot fix that right now. Maybe make a BSPO issue?

In any case I will see that I will deprecate all of CARO in the next days, seems we never officially did that.