Open DeniseSl22 opened 5 years ago
No problem with that perse. But since this actually forms a database of mapping files, with versioning, It would be better to mint our own IDs for these and have these registered in identifiers.org. We could still have DOIs for some of the important ones. Note that indentifiers.org is an ELIXIR recommended interoperability approach also because it offers additional features (like monitoring of parallel repositories) while DOIs are not (yet)
Ohh and even DOIs should be cited using identifiers.org...
Okay, so we should first create a scheme describing our data sets for identifiers.org .
@Chris-Evelo : I think the idea on minting our own ideas is a long term solution. For the short term, I think it would be better to just get started getting DOIs for our data. I've now also started uploading the interaction-mapping files to Figshare, so these also have a DOI. For the gene/protein mappings this is not the case (yet), but those first need to start working again.
And what about LinkSets (created for CyTargetLinker @mkutmon ) and the ones mentioned in #8 ? I think it's a good idea to make those also citable with DOIs, to make them more F and A (from FAIR).
I'm in the process of loading all linksets to NDEx which can provide DOIs. So should be fine.
@DeniseSl22 What is the problem with minting our own ID's it is not more complicated than the pathway IDs, is it?
The problem is that we do not have that in place right now (for datasets), while we can get them for free at FigShare/NDex etc.
Only the metabolomics mapping files and gene-variant mappings have a DOI (both from FigShare). I would like to see this for all mapping files, so people can cite it more easily!