Sample repository ontology

[mrpatrickwest]

Example sample: https://udco.tw.rpi.edu/vivo/display/n11018 Example repository: https://udco.tw.rpi.edu/vivo/display/n11180

[mrpatrickwest]

Problem with pulling in the sampling ontology is that it imports quite a few different ontologies

https://udco.tw.rpi.edu/vivo/listOntologies

All the ones on this page that do not have a namespace label.

Should we add labels and annotations so they show up looking good?

[mrpatrickwest]

https://udco.tw.rpi.edu/vivo/display/n10039 - one of the repositories entered into UDCO ... not that bad. Please review.

[mrpatrickwest]

@zednis @olyerickson @zhonghao1014 @am-e @xgmachina

please review

[mrpatrickwest]

New example sample: https://udco.tw.rpi.edu/vivo/display/n7813 New example repository: https://udco.tw.rpi.edu/vivo/display/n22533

[xgmachina]

How about mapping a few properties to be sub-property of dct:hasPart and others to dct:isPartOf

[mrpatrickwest]

YOu mean instead of using hasPart and isPartOf directly?

[zednis]

Are we using hasPart and isPartOf directly? I wasn't sure with dcosample:includesSample in the ontology.

dcterms:isPartOf instead of dcosample:includesSample? Or make dcosample:includesSample a sub-property of dcterms:isPartOf?

[mrpatrickwest]

Not sure why I did it this way, but SampleGroup -> hasPart -> Sample -> isPartOf -> SampleGroup

[mrpatrickwest]

seems that should just be includesSample, remove the domain, add as restriction to SampleRepository and SampleGroup

[mrpatrickwest]

and make includesSample and includesSampleGroup subproperties of hasPart, and the inverse a subproperty of isPartOf.

[zednis]

Not sure why I did it this way, but SampleGroup -> hasPart -> Sample -> isPartOf -> SampleGroup

This makes sense, though if SampleGroup is a samfl:SamplingCollection then it should use samfl:member.

samfl:SamplingCollection
  rdf:type owl:Class ;
  dct:description "Collection of features on which observations may be made, which are intended to be representative of one or more features that are not fully accessible for some reason. "^^xsd:string ;
  rdfs:comment "The member sampling features are intended to sample domain features, as indicated by one or more sampledFeature properties. If only one is present then this is the sampled feature of all the member sampling features, unless locally overridden."^^xsd:string ;
  rdfs:isDefinedBy <http://def.seegrid.csiro.au/ontology/om/sam-lite> ;
  rdfs:label "Collection of sampling features"@en ;
  rdfs:subClassOf [
      rdf:type owl:Restriction ;
      owl:minCardinality "1"^^xsd:nonNegativeInteger ;
      owl:onProperty samfl:member ;
    ] ;
.

[zednis]

and make includesSample and includesSampleGroup subproperties of hasPart, and the inverse a subproperty of isPartOf.

:+1:

[mrpatrickwest]

okay ... to summarize

includesSample -> sub property of -> hasPart (inerse sub property of isPartOf) includesSampleGroup -> sub property of -> hasPart (inerse sub property of isPartOf)

SampleGroup remove restrictions for hasPart since there's already a property samfl:member

Did I get that right?

everything is good.

[mrpatrickwest]

no inverse property for samfl:member so would add dcosample:memberOf to be the inverse property

[zednis]

includesSample -> sub property of -> hasPart (inerse sub property of isPartOf) includesSampleGroup -> sub property of -> hasPart (inerse sub property of isPartOf)

I think this is good; just to sanity check - this is the definition for dcterms:hasPart

A related resource that is included either physically or logically in the described resource.

Do we all agree this makes sense as a super-property of includesSample and includesSampleGroup?

• does it make sense to say a sample is included physically or logically in the repository?
• does it make sense to say a sample group is included physically or logically in the repository?

[mrpatrickwest]

yes, I think that makes sense.

[xgmachina]

yes, too on the use of dct:hasPart

[am-e]

I like the use of dct:hasPart and dct:isPartOf, does that take care of our previous discussion about the relationship between samples/sample groups and the repository?