Fix ORCID individuals import (#3257)

* extract orcids to import

* create rule for orcidio_import.owl

* import orcidio in uberon-edit

* change recipe to use filter

* make orcidio import custom in odk

---------

Co-authored-by: Anita Caron <[email protected]>

.gitignore

@@ -219,3 +219,4 @@ src/ontology/subsets/*.obo
 src/ontology/subsets/*.owl
 src/ontology/subsets/*.tsv
 src/ontology/imports/*_import.owl
+!src/ontology/imports/orcidio_import.owl

docs/odk-workflows/ManageDocumentation.md

@@ -43,5 +43,3 @@ The documentation is _not_ automatically updated from the Markdown, and needs to
 3. Just to double check, you can now navigate to your documentation pages (usually https://obophenotype.github.io/uberon/). 
    Just make sure you give GitHub 2-5 minutes to build the pages!
 
-
-

docs/odk-workflows/RepositoryFileStructure.md

@@ -28,7 +28,7 @@ These are the current imports in UBERON
 | ncbitaxon | http://purl.obolibrary.org/obo/ncbitaxon/subsets/taxslim.owl | None |
 | ncbitaxondisjoints | http://purl.obolibrary.org/obo/ncbitaxon/subsets/taxslim-disjoint-over-in-taxon.owl | None |
 | nbo | http://purl.obolibrary.org/obo/nbo.owl | None |
-| orcidio | https://w3id.org/orcidio/orcidio.owl | None |
+| orcidio | https://w3id.org/orcidio/orcidio.owl | custom |
 | omo | http://purl.obolibrary.org/obo/omo.owl | mirror |
 
 ## Components

src/ontology/Makefile

@@ -10,7 +10,7 @@
 # More information: https://github.com/INCATools/ontology-development-kit/
 
 # Fingerprint of the configuration file when this Makefile was last generated
-CONFIG_HASH=                16a64e727250fbc31f79a269800d3885cc70b4db1e6b69119f66415f340619fa
+CONFIG_HASH=                d9dd1f14c02c9157faa08c0c33d811609f31616c2917255ed4dca5559cb438ec
 
 
 # ----------------------------------------
@@ -408,6 +408,10 @@ $(IMPORTDIR)/%_import.owl: $(MIRRORDIR)/merged.owl $(IMPORTDIR)/%_terms_combined
 
 .PRECIOUS: $(IMPORTDIR)/%_import.owl
 
+## Module for ontology: orcidio
+
+$(IMPORTDIR)/orcidio_import.owl: $(MIRRORDIR)/merged.owl
+	echo "ERROR: You have configured your default module type to be custom; this behavior needs to be overwritten in uberon.Makefile!" && false
 ## Module for ontology: omo
 
 $(IMPORTDIR)/omo_import.owl: $(MIRRORDIR)/merged.owl $(IMPORTDIR)/omo_terms_combined.txt

src/ontology/catalog-v001.xml

@@ -51,6 +51,7 @@
       <uri name="http://purl.obolibrary.org/obo/uberon/components/mappings.owl" uri="components/mappings.owl"/>
       <uri name="http://purl.obolibrary.org/obo/uberon/components/vasculature_class.owl" uri="components/vasculature_class.owl"/>
       <uri name="http://purl.obolibrary.org/obo/uberon/imports/merged_import.owl" uri="imports/merged_import.owl"/>
+      <uri name="http://purl.obolibrary.org/obo/uberon/imports/orcidio_import.owl" uri="imports/orcidio_import.owl"/>
       <uri name="http://purl.obolibrary.org/obo/uberon/patterns/definitions.owl" uri="../patterns/definitions.owl"/>
       <uri name="http://purl.obolibrary.org/obo/wbbt.owl" uri="imports/local-wbbt.owl"/>
       <uri name="http://purl.obolibrary.org/obo/wbls.owl" uri="imports/local-wbls.owl"/>

src/ontology/uberon-edit.obo

@@ -63,6 +63,7 @@ import: http://purl.obolibrary.org/obo/uberon/components/in-subset.owl
 import: http://purl.obolibrary.org/obo/uberon/components/mappings.owl
 import: http://purl.obolibrary.org/obo/uberon/components/vasculature_class.owl
 import: http://purl.obolibrary.org/obo/uberon/imports/merged_import.owl
+import: http://purl.obolibrary.org/obo/uberon/imports/orcidio_import.owl
 import: http://purl.obolibrary.org/obo/uberon/patterns/definitions.owl
 ontology: uberon/core
 property_value: dc-contributor https://github.com/orgs/pato-ontology/teams/pato-community

src/ontology/uberon-odk.yaml

@@ -73,6 +73,7 @@ import_group:
       mirror_from: https://w3id.org/orcidio/orcidio.owl
       base_iris:
         - https://orcid.org/
+      module_type: custom
     - id: omo
       module_type: mirror
 subset_group:

src/ontology/uberon.Makefile

@@ -437,6 +437,20 @@ all_local_imports:
 
 endif
 
+# ----------------------------------------
+# EXTRACT ORCIDs INDIVIDUALS
+# ----------------------------------------
+
+# Extract ORCID from dcterms:contributor annotations
+$(IMPORTDIR)/orcidio_terms_combined.txt: $(SRC)
+	$(ROBOT) query -f csv -i $< --query ../sparql/orcids.sparql $@.tmp &&\
+	cat $@.tmp | sort | uniq >  $@
+
+# Create orcidio_import.owl to extract individuals excluded in the merged_import.owl
+$(IMPORTDIR)/orcidio_import.owl: $(MIRRORDIR)/orcidio.owl $(IMPORTDIR)/orcidio_terms_combined.txt
+	if [ $(IMP) = true ]; then $(ROBOT) filter -i $< -T $(IMPORTDIR)/orcidio_terms_combined.txt --trim false \
+	$(ANNOTATE_CONVERT_FILE); fi
+
 
 # ----------------------------------------
 # UNRELEASED PRODUCTS

src/sparql/orcids.sparql

@@ -0,0 +1,7 @@
+PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+prefix owl: <http://www.w3.org/2002/07/owl#>
+SELECT DISTINCT ?orcid
+WHERE {
+  ?term <http://purl.org/dc/terms/contributor> ?orcid . 
+  FILTER(isIRI(?term))
+}