Extend iedb to multiple semantic spaces

[bgyori]

Prefix

iedb

Explanation

The IEDB database is already registered as iedb. However, IEDB actually consists of identifiers for multiple semantic spaces:

• reference: these are internal IDs for sources of evidence like a publication, example: https://www.iedb.org/reference/1038233. This is what is currently assumed as the basis for the URI format in the Bioregistry.
• assay: a specific experimental assay serving as evidence for an entry. Example: https://www.iedb.org/assay/1691753
• antigen: provides information on a given antigen. Effectively operates as a provider for external IDs. Examples: https://www.iedb.org/antigen/UNIPROT:P62315 and https://www.iedb.org/antigen/ChEBI:143712
• epitope: part of a protein that is studied as an epitope. Example: https://www.iedb.org/epitope/117395
• receptor: receptors studied as a group. Example: https://www.iedb.org/receptor/47
• mhc: MHC restrictions of cells. Example: https://www.iedb.org/mhc/86

and currently the latter five are not appropriately registered. We could consider breaking up iedb into iedb.reference, iedb.assay, etc. without a parent iedb entry. This would follow the template of e.g., the pubchem.* entries in the Bioregistry.

Contributor ORCID

0000-0001-9439-5346

[bgyori]

A few issues related to this:

• Mappings for the current iedb entry:
  • miriam: this maps to what would become iedb.reference
  • prefixcommons: this maps to what would become iedb.epitope
  • fairsharing, integbio, re3data: these entries are all generic to all of IEDB so not clear what to do with them if we break up iedb into multiple semantic-space-specific entries. a) drop them, b) map them to one arbitrary new iedb entry or c) map them redundantly to all of the new iedb.* entries?
• Publications
  • There are 4 publications curated directly under iedb (plus additional ones listed under the mapped entries). These publications are all generic to all of IEDB so again not clear what happens if iedb is split up. a) do we add them to one arbitrary new iedb entry or b) redundantly add them to all entries? Looking at pubchem.* as an example in this regard I didn't find any patterns to follow, publication assignment to each entry there looks pretty arbitrary.

[bgyori]

Done in #1238