Create design patterns for neoplasm/cancer related terms

obophenotype / upheno

The Unified Phenotype Ontology (uPheno) integrates multiple phenotype ontologies into a unified cross-species phenotype ontology.

https://obophenotype.github.io/upheno/

Creative Commons Zero v1.0 Universal

76 stars 17 forks source link

Create design patterns for neoplasm/cancer related terms #145

Open cmungall opened 9 years ago

cmungall commented 9 years ago

See @drseb's question here: https://github.com/obophenotype/human-phenotype-ontology/issues/473#issuecomment-142003070

Briefly:

MP cancer terms typically refer to disposition
HP cancer terms refer to an actual material object and or process in a patient (although there are 2 related syns that pertain to disposition)

We would like to use NCIt. How? Need to be careful of ontological commitments. NCIt cancers are under 'finding'. We strictly segregate disease and phenotype. If we use the same class in both contexts (we will most likely switch NCIt classes into mondo) then trouble will ensue

lschriml commented 9 years ago

Why not use Disease Ontology for cancer terms ?

On Mon, Sep 21, 2015 at 11:08 AM, Chris Mungall notifications@github.com wrote:

See @drseb https://github.com/drseb's question here: obophenotype/human-phenotype-ontology#473 (comment) https://github.com/obophenotype/human-phenotype-ontology/issues/473#issuecomment-142003070

Briefly:

MP cancer terms typically refer to disposition

HP cancer terms refer to an actual material object and or process in a patient (although there are 2 related syns that pertain to disposition)

We would like to use NCIt. How? Need to be careful of ontological commitments. NCIt cancers are under 'finding'. We strictly segregate disease and phenotype. If we use the same class in both contexts (we will most likely switch NCIt classes into mondo) then trouble will ensue

— Reply to this email directly or view it on GitHub https://github.com/obophenotype/upheno/issues/145.

pnrobinson commented 9 years ago

Suggest: We should reference the disease entities to DO and the phenotypic feature (the actual tumour) to HP and we need to decide what semantics the NCI terms have, i.e., are they closer to HP (finding) or are they closer to DO (disease). -Peter

lschriml commented 9 years ago

If it is helpful, DO also contains tumor terms, our terms are aligned to NCIt terms.

Cheers, Lynn

mellybelly commented 9 years ago

NCIT is the authority here, prefer to get what we need there and all the rest of us pull from there? But, here we are discussing phenotypes not diseases.

pnrobinson commented 9 years ago

Yes, we need a strategy to (1) migrate and (2) rescue old annotations and maybe (3) wrap the terms in xrefs so as to make downstream software insulated from changes etc

cmungall commented 9 years ago

NCIt classes are axiomatized, links back to cells, anatomy etc (which we would rewire to uberon/cl)

I had also assumed that NCIt more complete for cancer. What is the long term plan for integration?

lschriml commented 9 years ago

DO is working directly with NCIt, to integrate their terms which are not yet in DO. We are working on a set of ~300 terms identified by NCI as potentially novel for DO.

drseb commented 9 years ago

Please note that every Neoplasm in NCIT is a subclass of Disease or Disorder, which is problematic for HP/MP.

drseb commented 8 years ago

this is slowly progressing. @cmungall @mellybelly do we have someone actively working on this?