search

OPMI / opmi

OPMI: Ontology of Precision Medicine and Investigation
https://obofoundry.org/ontology/opmi.html
Creative Commons Attribution 4.0 International
6 stars 3 forks source link

specimen barcode #8

Open PhiBabs935 opened 4 years ago

PhiBabs935 commented 4 years ago

Hi, I was thinking of potentially using OPMI's 'specimen barcode' in the Influenza Ontology, using it to replace FLU's version of 'specimen barcode' (as that term is not specific to influenza). But I wanted to ensure consistency first.

I noticed that the OPMI superclass 'barcode' is classified as a IAO:CRID. It appears to me that this isn't consistent with IAO's stance on CRIDs. According to an editor note on 'CRID', Universal Product Code barcodes are not CRIDs, as they are not centrally registered. In that case not every barcode is a CRID.

So I am not sure whether 'specimen barcodes' are CRIDs. Does a specimen barcode always have information about the CRID registry to which it belongs?

Currently, the FLU version of 'specimen barcode' is classified as a 'specimen ID', a term from OBIB. Specimen ID is a subclass of IAO:identifier.

yongqunh commented 4 years ago

This is a good point. However, we may first need to understand the meaning of "centrally registered" in centrally registered identifier symbol (CRID). The initial thought was that all specides barcodes should be somehow centrally registered. Otherwise, why we need to use them?

Also, is 'specimen barcode' as a 'specimen ID'? it may depend on how the ID or IAO:identifier is defined. CRID is an IAO:symbol. is 'specimen barcode' an IAO:symbol or IAO:identifier?

PhiBabs935 commented 4 years ago

You seem to be conflating 'CRID' with its proper part, 'CRID symbol'. OPMI:specimen barcode is a subclass of CRID, implying that it has a CRID symbol as a proper part. Of course, it could be re-classified as 'CRID symbol'. A specimen barcode would only be a CRID if it has both a CRID symbol and information content about the CRID registry it belongs to.

I didn't mean to imply that I had reasons to think that 'specimen barcodes' are not centrally registered. I just didn't know for sure. Of course, a specimen barcode can be centrally registered even if not all barcodes are. Though, if that is the case 'specimen barcode', but not 'barcode', would be a CRID symbol. If specimen barcodes are all centrally registered, 'specimen barcode' could perhaps be classified as a barcode, but given axioms that ensure it is an inferred subclass of CRID symbol.

The Influenza Ontology originally classified its own specimen barcode term as a CRID symbol, but OBIB imported it as an identifier instead. There seems to be some issues concerning classifying terms as 'identifier' or 'symbol'. Recently I raised an issue about the relationship between OBIB's 'specimen ID' and OBI's 'specimen identifier', which is classified as a CRID symbol. See https://github.com/obi-ontology/obi/issues/1246. That discussion led to them question whether 'specimen identifiers' have to be CRID symbols.

So, that said, I think that whether 'specimen barcode' should go under 'identifier' or 'symbol' is something that is up for debate. And that is why I wanted to run it by you, given that I was thinking of using the OPMI 'specimen barcode'. It seems like the direction that is suggested in the OBI issue that I linked to is to make 'specimen barcode' an identifier. But nothing is set in stone at this point.