search

phenopackets / phenopacket-reference-implementation

Reference java implementation for phenopackets
BSD 3-Clause "New" or "Revised" License
6 stars 9 forks source link

Add metadata for associations #69

Open cmungall opened 8 years ago

cmungall commented 8 years ago

Associations should have at least the person/group that created or contributed to the association (mapped to dc:creator and dc:contributor) plus creation timestamp.

balhoff commented 8 years ago

Should we support the creator being possibly either a string or an IRI? e.g. text name vs. ORCID

I guess having the value as an Entity would work for this—is it a problem for an Entity to have a null id?

cmungall commented 8 years ago

I had been thinking of the Entity being an entity in the main 'domain of discourse'. But this gets awkward if we have self-phenotyping humans.

I think Entity makes the most sense. We should then force an ID/IRI, no nulls.

On 6 Jul 2016, at 11:29, Jim Balhoff wrote:

Should we support the creator being possibly either a string or an IRI? e.g. text name vs. ORCID

I guess having the value as an Entity would work for this—is it a problem for an Entity to have a null id?

You are receiving this because you authored the thread. Reply to this email directly or view it on GitHub: https://github.com/phenopackets/phenopacket-reference-implementation/issues/69#issuecomment-230862627

balhoff commented 8 years ago

How fine-grained should this be? We could add contributor and date fields to Association, and maybe also to PhenoPacket. But should it support multiple contributors in each location? If multiple, would each have a separate date?

balhoff commented 8 years ago

We need to implement some form of this to pass contributor and date through PXF when going from HPO Phenote to Noctua OWL.