mixs icon indicating copy to clipboard operation
mixs copied to clipboard
verified publisher icon GenomicsStandardsConsortium

PURL for RDF/XML OWL for the latest production release of MIxS for use by OLS

Open Woolly-at-EBI opened this issue 6 months ago • 9 comments

Aim: have a PURL to an OWL file that can be uploaded into OLS (request by PRIDE's Yasset Perez [email protected] ).

Considerations

  1. Henriette Harmse [email protected] of OLS provided some guidance: How-are-ontologies-added-to-OLS - .owl file in RDF/XML ( and not in OWL/XML is fine) - purl constant across releases. (for the latest release)
  2. Do we need to have a simple RDF, that remove some of the existing complexity? (raised by Mark, we hope that we don't)
  3. What do we with the fact that there are multiple contextual definitions for some terms? Do we provided them all?
  4. would the preferredPrefix be: mixs? (then aligned with the BioPortal albeit uppercase https://bioportal.bioontology.org/ontologies/MIXS)
  5. If a description is needed, is it fine to identically use "An OWL representation of the Minimum Information for any (x) Standard (MIxS), managed by the Genomic Standards Consortium."
  6. How/who would generated this? I am happy to work with someone to achieve this.

When ready, Peter to raise issue https://github.com/EBISPOT/ols4/issues to have incorporated into OLS.

Woolly-at-EBI avatar Jun 24 '25 17:06 Woolly-at-EBI

Pushing ahead with this, as increasing need from EBI users at least and precedence already set with the presence in bioportal.

  1. No action needed, as already in OWL https://github.com/GenomicsStandardsConsortium/mixs/blob/main/project/owl/mixs.owl.ttl - thanks Mark
  2. Irrelevant
  3. Ignoring for now
  4. Going with "MIXS" as that is what is done in the rest of OLS and is . I don't know where my lower case aspect came from, besides my normal bias.
  5. No objection from anyone with the definition above, Can change if folk have antibodies.
  6. Talked with Chris M. previous and Lynn at GSC meeting about this

Woolly-at-EBI avatar Aug 05 '25 13:08 Woolly-at-EBI

Oh bother, been filling out the useful https://github.com/EBISPOT/ols4/blob/dev/New%20OLS%20ontology%20request.xlsx Houston, I think that we have a problem, the MIXs is all indexed off the structured name rather than the identifier I was anticipating and is more common in ontologies. Is it straightforward to generate an OWL version that indexes that uses the id e.g MIXS:0000013, rather than MIXS:env_local_scale?

Example snippet

prompt% grep -A 5  env_local_scale ~/Downloads/mixs.owl.ttl | less
MIXS:env_local_scale a owl:DatatypeProperty ;
    rdfs:label "env_local_scale" ;
    dcterms:title "local environmental context" ;
    schema:keywords "context",
        "environmental" ;
    skos:definition "Report the entity or entities which are in the sample or specimen s local vicinity and which you believe have significant causal influences on your sample or specimen. We recommend using EnvO terms which are of smaller spatial grain than your entry for env_broad_scale. Terms, such as anatomical sites, from other OBO Library ontologies which interoperate with EnvO (e.g. UBERON) are accepted in this field. EnvO documentation about how to use the field: https://github.com/EnvironmentOntology/envo/wiki/Using-ENVO-with-MIxS" ;
    skos:inScheme <https://w3id.org/mixs> ;

Further commentary

This may be fine for some of the implementers that just use the structured field names, but is less suited for such as ENA that use the title name, and it will also make MIxS more robust the structured field name name changes.

Woolly-at-EBI avatar Aug 05 '25 13:08 Woolly-at-EBI

MIXs is all indexed off the structured name rather than the identifier I was anticipating and is more common in ontologies

rebuilding the OWL after this is merged should fix it:

#1045

Note that of course there are OWL classes generated for things that are not MIxS terms - e.g. permissible values in enums

cmungall avatar Aug 05 '25 14:08 cmungall

Thanks Chris and Mark for the continuing education and setting in a fix. Looking forwards to the TWG at the hour.

Woolly-at-EBI avatar Aug 05 '25 14:08 Woolly-at-EBI

I have drafted what I can for the information to send to OLS. Please can someone have a butchers and hack where appropriate ? Thank you in advance.

Google sheet MIxS "ontology" details for OLS

Woolly-at-EBI avatar Aug 06 '25 16:08 Woolly-at-EBI

I have submitted the MIxS ontology details to OLS + attached the mixs.owl.ttl with the MIxS IDs. The request has been received by OLS, but they are awaiting a PURL for the mixs.owl.ttl and a few other questions. https://github.com/EBISPOT/ols4/issues/951 EBI colleagues have dependencies that needs this in OLS ASAP, so I am trying to push this along as fast as I can, in between others responsibilites, I think that this would help with interoperability of MIxS.

  1. For mixs.owl.ttl , how does this get updated in https://github.com/GenomicsStandardsConsortium/mixs/blob/main/project/owl/mixs.owl.ttl ? Is it updated by a trigger, else, do I just commit the ttl that contains the MIxS id?
  2. would this be the best place for a PURL for the owl? https://w3id.org/mixs/mixs.owl.ttl and redirects too https://github.com/GenomicsStandardsConsortium/mixs/blob/main/project/owl/mixs.owl.ttl If so please can this be created?
  3. who is the best technical contact with MIxS going forwards, as it is no longer Ramona. Would it be Montana. Who are contributors? Chris H., Lynn Schriml, Chris M, Mark M., Sujay, Montana? I (Peter is currently the submitter)

Woolly-at-EBI avatar Aug 11 '25 13:08 Woolly-at-EBI

For 2, yes, use https://w3id.org/mixs/mixs.owl.ttl

On Mon, Aug 11, 2025 at 6:52 AM Peter Woollard @.***> wrote:

Woolly-at-EBI left a comment (GenomicsStandardsConsortium/mixs#1006) https://github.com/GenomicsStandardsConsortium/mixs/issues/1006#issuecomment-3174947396

I have submitted the MIxS ontology details to OLS + attached the mixs.owl.ttl with the MIxS IDs. The request has been received by OLS, but they are awaiting a PURL for the mixs.owl.ttl and a few other questions. EBISPOT/ols4#951 https://github.com/EBISPOT/ols4/issues/951 EBI colleagues have dependencies that needs this in OLS ASAP, so I am trying to push this along as fast as I can, in between others responsibilites, I think that this would help with interoperability of MIxS.

  1. For mixs.owl.ttl , how does this get updated in https://github.com/GenomicsStandardsConsortium/mixs/blob/main/project/owl/mixs.owl.ttl ? Is it updated by a trigger, else, do I just commit the ttl that contains the MIxS id?
  2. would this be the best place for a PURL for the owl? https://w3id.org/mixs/mixs.owl.ttl and redirects too https://github.com/GenomicsStandardsConsortium/mixs/blob/main/project/owl/mixs.owl.ttl If so please can this be created?
  3. who is the best technical contact with MIxS going forwards, as it is no longer Ramona. Would it be Montana. Who are contributors? Chris H., Lynn Schriml, Chris M, Mark M., Sujay, Montana? I (Peter is currently the submitter)

— Reply to this email directly, view it on GitHub https://github.com/GenomicsStandardsConsortium/mixs/issues/1006#issuecomment-3174947396, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAAMMOPUMWDNNGWS74YXFKL3NCNZJAVCNFSM6AAAAACAA6NIZSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTCNZUHE2DOMZZGY . You are receiving this because you commented.Message ID: @.***>

cmungall avatar Aug 11 '25 23:08 cmungall

  1. who is the best technical contact with MIxS going forwards, as it is no longer Ramona. Would it be Montana. Who are contributors? Chris H., Lynn Schriml, Chris M, Mark M., Sujay, Montana? I (Peter is currently the submitter)

I'm not "technical" but you can put me down as the contact if no-one else wants to be added!

only1chunts avatar Aug 12 '25 07:08 only1chunts

In parallel, I have been pushing ahead with aspects on the OLS side, so that the elapsed time will be less. They provided some feedback on the OWL format we generate, and their feedback looks valid to me. Please can one of you who understands this deeper agree or disagree? And if needed can the linkml to OWL generator be fixed?

  • matentzn left a comment (EBISPOT/ols4#951) As a sideline lurker I want to note that your use of owl:Class is wrong; it should be

owl:Class NOT owl:Class which looks like this: image.png (view on web)

  • In our current OWL file:
<https://w3id.org/mixs/TaxIdentEnum#16S+rRNA+gene> **a <owl:Class>** ;
    rdfs:label "16S rRNA gene" ;
    dcterms:isPartOf MIXS:TaxIdentEnum ;
    rdfs:subClassOf linkml:PermissibleValue .
  • To be example?
<https://w3id.org/mixs/TaxIdentEnum#16S+rRNA+gene> **a owl:Class** ;
    rdfs:label "16S rRNA gene" ;
    dcterms:isPartOf MIXS:TaxIdentEnum ;
    rdfs:subClassOf linkml:PermissibleValue .

I have been improving my knowledge of this all:

  • information youtube article on TTL
  • https://www.w3.org/TR/owl2-syntax/
  • https://robot.obolibrary.org/convert.html
  • https://linkml.io/linkml/generators/owl.html
  • https://www.w3.org/TR/turtle/
  • https://www.w3.org/TR/owl-ref/ <-- one of the most useful

Woolly-at-EBI avatar Sep 02 '25 15:09 Woolly-at-EBI