create ado.md

[BideZ]

Hello, everyone. This is the metadata record created for ADO.
This ADO ontology refers to the one submitted in #1806

Thank you very much!

[matentzn]

• Blocked by non-resolvable version IRI still! I will add the activity status right after this is addressed.

[matentzn]

let me know when the version IRI is resolvable, I will then merge. :)

[BideZ]

Sorry I get a little confused here. Could you please specify how/where the version IRI is non-resolvable? :( The version IRI https://bio.scai.fraunhofer.de/ontology/ADO/2.0.0 should work in theory.

[matentzn]

Yes exactly! If it resolves not just in theory, but also in practice, you are sorted!

[matentzn]

wget https://bio.scai.fraunhofer.de/ontology/ADO/2.0.0
--2022-06-17 12:09:20--  https://bio.scai.fraunhofer.de/ontology/ADO/2.0.0
Resolving bio.scai.fraunhofer.de (bio.scai.fraunhofer.de)... 193.175.167.101
Connecting to bio.scai.fraunhofer.de (bio.scai.fraunhofer.de)|193.175.167.101|:443... connected.
ERROR: no certificate subject alternative name matches
	requested host name ‘bio.scai.fraunhofer.de’.
To connect to bio.scai.fraunhofer.de insecurely, use `--no-check-certificate'.

[cthoyt]

Would be nice to finish #1841 before admitting new ontologies

[BideZ]

I will discuss this with our team on Monday. Thank you for your feedback.

[BideZ]

@matentzn I have discussed this with our team. We currently have no persistent URL for ADO yet, and therefore the IRI remains non-resolvable. This holds for EPIO ontology during the submission.

[BideZ]

@cthoyt We are now proposing the CONTRIBUTING file. It might take some time, though. We will update the metadata file after this.

[matentzn]

@BideZ unfortunately Version IRIs must be resolvable - for EPIO, a glitch in the technical infrastructure prevented us from checking this.

You have two alternatives - either you adopt OBO style version IRIs, or you get your own version IRIs to resolve. One of the two is mandatory: https://obofoundry.org/principles/fp-004-versioning.html

In addition to an IRI specifying the current release (see Principle 3), each official release MUST have a unique version IRI that resolves to the specific ontology artifact indicated.

[BideZ]

@matentzn I have changed the version IRI according to the versioning principle. https://obofoundry.org/principles/fp-004-versioning.html
Could you please check if the IRI is resolvable now?

[matentzn]

Hey @BideZ unfortunately not. You will have to talk to your Fraunhofer technical team. Your certificate does not work!

Try putting your new version IRI in your Chrome and see what chrome tell you:

https://bio.scai.fraunhofer.de/ontology/ADO/2022-02-21/ADO.owl

Or, if you have a Mac or Linux, try

wget https://bio.scai.fraunhofer.de/ontology/ADO/2022-02-21/ADO.owl

It will complain about a missing certiciate.

[BideZ]

@matentzn We are wondering if we could acquire a PURL from you since Fraunhofer do not host any ontology itself, and we think that is the main problem causing this issue.
We thought this is how it works for the version IRIs:

1. We propose an IRI for our ontology
2. Obtain a PURL
3. Link the PURL to the ontology in Github

Please correct me if it works in other ways

[matentzn]

If you:

• Change your ontology IRI to http://purl.obolibrary.org/obo/ado.owl
• Change your ontology version IRI to http://purl.obolibrary.org/obo/ado/releases/2022-06-11/ado.owl

You can use the OBO PURL system for managing your releases! Else you will have to deal with the redirects yourself at Fraunhofer.

When your registration is complete, you will be allocated the above PURLs for your ontology. But they will have to be used in your ontology as well!

[BideZ]

@matentzn Thank you for the reply. We have changed the ontology IRI as well as the version IRI to the aforementioned IRIs. This has been applied to all the original classes inside of our ontology too.

[matentzn]

@BideZ your ontology and version IRIs are ok now, but all your entity IRIs are wrong.

oboInOwl

oboinowl: is broken. These are both very wrong:
xmlns:oboInOwl1="http://purl.obolibrary.org/obo/ado.owl#oboInOwl:"
xmlns:oboInOWL="http://www.geneontology.org/formats/oboInOWL#"

they should be:
xmlns:oboInOwl="http://www.geneontology.org/formats/oboInOwl#"

Make sure you only ever use oboInOwl.

General URI scheme

This is the only official term scheme for OBO:

http://purl.obolibrary.org/obo/ADO_0000001

Nothing else is standard.. So this:

http://purl.obolibrary.org/obo/ado.owl#CommentOnDefinition

should ideally be

http://purl.obolibrary.org/obo/ADO_0000001

Similarly http://purl.obolibrary.org/obo/ado.owl#OWLObjectProperty_00000000000000000000

Should be:

http://purl.obolibrary.org/obo/ADO_0000000

I guess it would be ok to have it:

http://purl.obolibrary.org/obo/ADO_00000000000000000000

But if you do that, all your ADO ids should have 00000000000000000000 characters. I don't recommend it..

Wrong definition property:

http://purl.obolibrary.org/obo/ado.owl#oboInOwl:hasDefinition 

is broken, should be:

http://purl.obolibrary.org/obo/IAO_0000115

[matentzn]

Also, in order for this:

http://purl.obolibrary.org/obo/ado/releases/2022-06-11/ado.owl

To be valid, you will have to create a release on Github that has v2022-06-11 as a tag. This is necessary to redirect from the OBO purl system to your version. Note that if you use this IRI, then 2022-06-11 becomes your official versioning scheme - not 2.0.0!

[BideZ]

@matentzn Thank you for the feedback. We have changed the class IRIs to the recommended format except the two annotation properties whose names would get lost. The release was created and the directory name was reformated according to the new versioning scheme. Please have a look.

[matentzn]

Much better, we are getting there :)

• http://purl.obolibrary.org/obo/ADO#CommentOnDefinition is still non-standard, see above! Should be: http://purl.obolibrary.org/obo/ADO_123567. If you cant do that right now, it is acceptable to use http://purl.obolibrary.org/obo/ado#CommentOnDefinition (lower case ado). OBO does not really allow this, but many ontologies do it, so I would wave it through now. But: numeric ids are preferred!
• oboInOwl:hasDefinition should be http://purl.obolibrary.org/obo/IAO_0000115
• A few Copy/paste errors of this type: http://purl.obolibrary.org/obo/ado.owlpurl.obolibrary.org/obo/ADO_0000013

[BideZ]

@matentzn Thanks for the feedback. I have addressed the issues and it is available now through our latest push.

[matentzn]

@BideZ Good Job!! Sorry it was a bit tough (sorry), but you made it: https://obofoundry.org/obo-nor.github.io/dashboard/ado/dashboard.html

This is now ready for merge.

[matentzn]

@BideZ I will merge this now, but it would be great if you could make another PR to complete the list of dependencies. You seem to be reusing terms from many ontologies, including GECKO, NCBITaxon and others - it would be good to list them here.