What are GC_IDs?

[cthoyt]

Most terms have an xref to a namespace with a prefix GC_ID. Is anyone familiar with what that is or what it abbreviates?

[jamesaoverton]

I have a partial answer. The ncbitacon.owl file is a direct translation of the taxdmp.zip file available here: https://ftp.ncbi.nih.gov/pub/taxonomy/. In that directory is a taxdmp_readme.txt that explains the various fields. "GC" is their abbreviation for "genetic code", and points to a gencode.dmp file that we do not translate. Official NCBI Taxonomy pages include a "Genetic code" field with a link, e.g. https://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?mode=Info&id=9606&lvl=3&lin=f&keep=1&srchmode=1&unlock. That's as much as I know.

[cthoyt]

Thanks @jamesaoverton, that's much appreciated. It's unbelievable how many nomenclatures the NCBI has generated...

[cmungall]

FWIW UMLS doesn't translate this either https://www.nlm.nih.gov/research/umls/sourcereleasedocs/current/NCBI/sourcerepresentation.html

I suggest

1. Register something like NCBI.gc with identifiers.org / n2t.net
2. Point this at URLs like https://www.ncbi.nlm.nih.gov/Taxonomy/Utils/wprintgc.cgi#SG2 or https://www.ncbi.nlm.nih.gov/Taxonomy/Utils/wprintgc.cgi?chapter=tgencodes#SG2
3. have an annotation/xref pointing to this
4. (stretch) have some kind of ontological rendering of gencode.dmp (btw, did the file move? I don't see it). E.g
   • taxon has-part (nuclear genome and has-part some translation system GC_ID)
   • GC_ID a SP:codon, label "ATG", starts-with some (adenine and followed-by thymine and ends-with guanine) encodes chebi:methionine]
   • this injects a bunch of blank nodes into the ontology with no real priority use case and would be for the sake of ontological completeness, so YMMV....

[cthoyt]

FYI: This has been registered in the Bioregistry at http://bioregistry.io/registry/gc

[cthoyt]

@cmungall @jamesaoverton I have got a working prototype of ontologizing GC in biopragmatics/pyobo#250, but I'm wondering if an approach more like TAXRANK where I hand-curate the list, then xref back to the NCBI vocabulary is best. I also went a bit off script in the PyOBO PR and added some non-isomorphic stuff, like defining an external "has genetic code" type definition as well as added a few mid-level classes to categorize nuclear, mitochondrial, and plastid genomes. Given I did that, maybe it makes sense to go the TAXRANK route of making this a manually-curated hub that could also be a place where we could add some of the ontologizations chris mentioned above