search

ncbo / ncbo_annotator

To automatically process a piece of data text to annotate it with relevant ontology concepts and return the annotations.
http://bioportal.bioontology.org/annotator
Other
19 stars 9 forks source link

Semantic types aren't universally recognized #5

Open mdorf opened 5 years ago

mdorf commented 5 years ago

Per user report:

I have been experimenting with the annotator found here: https://bioportal.bioontology.org/annotator

Settings: Text entered into annotator: invasive mammary carcinoma Select ontology: National Cancer Institute Thesaurus (NCIT) Select UMLS semantic type: neoplastic process

Result: no annotations found

I am not understanding why the result is 'No annotations found'. The term 'invasive mammary carcinoma' does exist in the NCIT, with semantic_type 'Neoplastic process'.

https://bioportal.bioontology.org/ontologies/NCIT?p=classes&conceptid=http%3A%2F%2Fncicb.nci.nih.gov%2Fxml%2Fowl%2FEVS%2FThesaurus.owl%23C9245

mdorf commented 5 years ago

Our logic for annotating semantic types looks for a property of the class that contains the semantic type code (T191 in this case). Once found, the value is used to set a custom top-level property of the class called “semanticType”. The value of the “semanticType” property is then used when we process classes for the Annotator repository. Looking at the class you provided, there is no property in the class that references the semantic type code. There is one that states the name of the semantic type, “Neoplastic Process”, but that isn’t sufficient for us to classify this as a class belonging to the T191 semantic type:

http://data.bioontology.org/ontologies/NCIT/classes/http%3A%2F%2Fncicb.nci.nih.gov%2Fxml%2Fowl%2FEVS%2FThesaurus.owl%23C9245?display=all&no_links=true&no_context=true

Here is a sample of a different class from SNOMEDCT that does set the semantic type property and thus appears in your annotation query (if you remove the NCIT ontology filter):

http://data.bioontology.org/ontologies/SNOMEDCT/classes/http%3A%2F%2Fpurl.bioontology.org%2Fontology%2FSNOMEDCT%2F68453008?display=all&no_links=true&no_context=true

graybeal commented 3 years ago

I'm making this a "won't fix", as we are not doing anything wrong per Misha's analysis.