In ontobio.model.nlp (https://github.com/biolink/ontobio/blob/master/ontobio/model/nlp.py#L19) the field category is always empty, since SciGraph appears to the return the field as categories. I can't find precisely where or when in the SciGraph commit history the field was named categories, but you can see from this query that the returned field name is currently categories:
curl -X POST "https://scigraph-ontology.monarchinitiative.org/scigraph/annotations/entities" -H "accept: application/json" -H "content-type: application/x-www-form-urlencoded" -d "content=male&minLength=4&longestOnly=false&includeAbbrev=false&includeAcronym=false&includeNumbers=false"
The result being:
[
{
"token": {
"id": "UBERON:0003101",
"categories": [
"anatomical entity"
],
"terms": [
"male organism"
]
},
"start": 0,
"end": 4
},
{
"token": {
"id": "PATO:0000384",
"categories": [
"quality"
],
"terms": [
"male"
]
},
"start": 0,
"end": 4
},
{
"token": {
"id": "WBbt:0007850",
"categories": [
"anatomical entity"
],
"terms": [
"male"
]
},
"start": 0,
"end": 4
}
]If this is in fact a mislabeled field, both the field in ontobio.model.nlp.Token and the field in biolink.datamodel.serializers (https://github.com/biolink/biolink-api/blob/master/biolink/datamodel/serializers.py#L336) will need to be corrected.
(This issue comes from investigating monarch-initiative/biolink-api#387.)
In
ontobio.model.nlp(https://github.com/biolink/ontobio/blob/master/ontobio/model/nlp.py#L19) the fieldcategoryis always empty, since SciGraph appears to the return the field ascategories. I can't find precisely where or when in the SciGraph commit history the field was namedcategories, but you can see from this query that the returned field name is currentlycategories:The result being:
[ { "token": { "id": "UBERON:0003101", "categories": [ "anatomical entity" ], "terms": [ "male organism" ] }, "start": 0, "end": 4 }, { "token": { "id": "PATO:0000384", "categories": [ "quality" ], "terms": [ "male" ] }, "start": 0, "end": 4 }, { "token": { "id": "WBbt:0007850", "categories": [ "anatomical entity" ], "terms": [ "male" ] }, "start": 0, "end": 4 } ]If this is in fact a mislabeled field, both the field in
ontobio.model.nlp.Tokenand the field inbiolink.datamodel.serializers(https://github.com/biolink/biolink-api/blob/master/biolink/datamodel/serializers.py#L336) will need to be corrected.(This issue comes from investigating monarch-initiative/biolink-api#387.)