search

edamontology / edam-bioimaging

Ontology of bioimage informatics operations, topics, types of data, and data formats. EDAM-Bioimaging is an extension of the EDAM ontology (edamontology.org) dedicated to bioimaging data analysis, and developed in an open collaboration including partners from NEUBIAS (neubias.org), COMULIS (comulis.eu), and ELIXIR (elixir-europe.org)..
https://webprotege.stanford.edu/#projects/2ce704bf-83ed-4d2e-985f-84c4841fac71/edit/Classes
Creative Commons Attribution Share Alike 4.0 International
8 stars 8 forks source link

Please get Machine Learning terms from Imaging to the Main Branch #33

Open rsmsheriff opened 11 months ago

rsmsheriff commented 11 months ago

We would like to use EDAM to annotate AI ML models in BioModels. See our recent manuscript. https://www.biorxiv.org/content/10.1101/2023.05.22.540599v1

Could you please do that soon? We have been waiting for this since last year.
We would like to have stable identifiers for the ML terms.

Many thanks in advance.

matuskalas commented 10 months ago

Dear @rsmsheriff , thank you so much for the prod 🙏🏽

I completely understand your frustration, and I'm even more frustrated than you, about the fact that this couldn't have happened much faster. The main step is the assignment of persistent IDs to the EDAM Bioimaging (incl. ML) concepts, which has been waiting for the next major release of the mainline EDAM, in order to be doable without a major mess-up (technological debt). And the next major release of EDAM has been being delayed over and over, uncovering ever new inconsistencies from the past🙁

Luckily, just a couple of weeks before you posted this issue, I came up with a dirty hack that will enable the PID assignment - and with that unlimited usage in various resources including BioModels, Bio.tools, TeSS, Galaxy, etc. - without a major technological debt for the future.

In the best case, I might be able to fix this already now during the BioHackathon, 🤞🏽