monarch-initiative / monarch-legacy

Monarch web application and API
BSD 3-Clause "New" or "Revised" License
42 stars 37 forks source link

/gene/ json results need tweaking #306

Closed cmungall closed 9 years ago

cmungall commented 10 years ago

E.g. Note some of these will require further discussion

http://127.0.0.1:8282/gene/NCBIGene:388552.json

General

cmungall commented 10 years ago

Note we have

    "source": {
        "id": "nlx_151671-2",
        "label": "ClinVar, via Office of Rare Diseases (9435)"
    }

And

"source": {

    "id": "nlx_151671-2",
    "label": "ClinVar, via NCBI curation"

}

The relationship between id and label is functional.

The via part should probably be modeled in the relationship between data and source (it is already?)

mellybelly commented 10 years ago

@cmungall I would use OBI or ERO for techniques for the interactions. We'll need these for expression and phentoyping in some cases too. We should pull in a slice of these two ontologies - assays and techniques. @mbrush can you help with this?

For summary, maybe one of these: http://purl.org/dc/terms/description http://www.w3.org/2004/02/skos/core#definition Probably prefer the former.

mbrush commented 9 years ago

@mellybelly the techniques in the BioGrid interaction data map to the PSI-MI 'ontology' (http://purl.obolibrary.org/obo/mi.owl), which is actually a pretty rich vocabulary. We should consider if we would use the IRIs from this ontology, or try to get terms into OBI/ERO with mappings to PSI-MI terms. As for any additional techniques needed, I can work to get them into OBI and/or ERO as needed.

BioGrid techniques are here: http://wiki.thebiogrid.org/doku.php/experimental_systems Earlier JIRA ticket about this is here: https://support.crbs.ucsd.edu/browse/NIF-11813

kshefchek commented 9 years ago

No longer relevant but we should create a new issue for exporting data as JSON.