Performing extract using ROBOT excludes GCIs

[cmungall]

Reference (mostly for me): https://github.com/geneontology/go-ontology/issues/13725

We're stuck using owltools for module extraction in GO for bio-chebi for now. It's weird as both owltools and robot use SLME under the hood here.

Mostly for my own reference:

ROBOT_JAVA_ARGS=-Xmx8G robot extract -i bio-chebi.owl -T imports/chebi_terms_combined.txt --method BOT -O http://purl.obolibrary.org/obo/go/imports/chebi_import.owl annotate -V http://purl.obolibrary.org/obo/go/releases/`date +%Y-%m-%d`/imports/chebi_import.owl -o imports/chebi_import.owl

this should produce 1000s of GCIs, as seen here: http://purl.obolibrary.org/obo/go/imports/chebi_import.owl

However, robot produces a single lonely GCI:

    <!-- 
    ///////////////////////////////////////////////////////////////////////////////////////
    //
    // General axioms
    //
    ///////////////////////////////////////////////////////////////////////////////////////
     -->

    <owl:Restriction>
        <owl:onProperty rdf:resource="http://purl.obolibrary.org/obo/RO_0000087"/>
        <owl:someValuesFrom rdf:resource="http://purl.obolibrary.org/obo/CHEBI_37527"/>
        <rdfs:subClassOf>
            <owl:Restriction>
                <owl:onProperty rdf:resource="http://purl.obolibrary.org/obo/GOCHEREL_0000004"/>
                <owl:someValuesFrom rdf:resource="http://purl.obolibrary.org/obo/CHEBI_37527"/>
            </owl:Restriction>
        </rdfs:subClassOf>
    </owl:Restriction>

Assigning to myself to make a better test case

[cmungall]

example of something we would hope to find:

    <owl:Restriction>
        <owl:onProperty rdf:resource="http://purl.obolibrary.org/obo/RO_0002345"/>
        <owl:someValuesFrom rdf:resource="http://purl.obolibrary.org/obo/CHEBI_29067"/>
        <owl:equivalentClass>
            <owl:Restriction>
                <owl:onProperty rdf:resource="http://purl.obolibrary.org/obo/RO_0002345"/>
                <owl:someValuesFrom rdf:resource="http://purl.obolibrary.org/obo/CHEBI_33575"/>
            </owl:Restriction>
        </owl:equivalentClass>
    </owl:Restriction>

[cmungall]

It's not even totally clear to me this is a robot issue, it may be something to do with how the seed classes are provided, investigating...

[balhoff]

@cmungall I tried your robot command above, and got a whole bunch of GCIs. Is this still a problem?

[cmungall]

are you doing the exact same command? I'm still failing to get the GCIs.

in fact now I get:

   <!-- 
    ///////////////////////////////////////////////////////////////////////////////////////
    //
    // General axioms
    //
    ///////////////////////////////////////////////////////////////////////////////////////
     -->

    <owl:Restriction>
        <owl:onProperty rdf:resource="http://purl.obolibrary.org/obo/RO_0000087"/>
        <owl:someValuesFrom rdf:resource="http://purl.obolibrary.org/obo/CHEBI_37527"/>
        <rdfs:subClassOf rdf:nodeID="_:genid8084"/>
    </owl:Restriction>

[balhoff]

Yes, exact same command. I just tried it again with a clean checkout, and it worked. Have you updated recently? I made a recent commit which changes how seed.txt is generated. That may be affecting it.

[beckyjackson]

Are you still having problems with this?

[ramonawalls]

@cmungall are you using GCI to mean the same thing as GCA, or is this something new I have never heard of?