thoughts about ontology connections?

[vjcitn]

I have been looking through the capacity of Cell Line Ontology and ChEBI to
give formal definitions of tokens in the cell_line and compounds components.
I think it can be done, although of course there will be gaps. Are you looking into
this? I am about to push github.com/vjcitn/RxGeno that demonstrates some of
the issues.

[vjcitn]

I see that you are using cellosaurus as the base ontology for cell lines. I use
the API to extract data, and only the internal IDs seem to be returned. Can one
retrieve the cellosaurus tags too?

[p-smirnov]

@vjcitn I don't believe we have this data exposed but we definitely can implement an endpoint to recover it from PharamcoDB.

We have tried to work with the CLO several years ago (!~2013) and had mixed results. However the CLO seems to have been under active development since then and I think its definitely worthwhile to try to map our database to the formal ontologies.

[p-smirnov]

@assefamaru Can you comment on the retrieving cellosaurus through the API

[assefamaru]

@vjcitn @p-smirnov PharmacoDB API does not fully parse/expose cellosaurus data, but an API for cellosaurus has already been implemented in a separate package: https://github.com/assefamaru/cellosaurus-api. Data is extracted from the structured text files provided by Cellosaurus: https://github.com/calipho-sib/cellosaurus/blob/master/cellosaurus.txt.