From 38f16f70c0e5c0cae4cbf253f2482781be408185 Mon Sep 17 00:00:00 2001
From: Nico Matentzoglu <nicolas.matentzoglu@gmail.com>
Date: Wed, 8 Jun 2022 21:40:02 +0300
Subject: [PATCH] Experimental pipeline for extracting classification
 mismatches

Experimental pipeline for extracting classification mismatches with SPARQL
---
 src/ontology/mondo-ingest.Makefile            | 23 +++++++++++++
 .../mondo-ordo-unsupported-subclass.sparql    | 34 +++++++++++++++++++
 2 files changed, 57 insertions(+)
 create mode 100644 src/sparql/mondo-ordo-unsupported-subclass.sparql

diff --git a/src/ontology/mondo-ingest.Makefile b/src/ontology/mondo-ingest.Makefile
index a9017aaa5..5cc4caf5f 100644
--- a/src/ontology/mondo-ingest.Makefile
+++ b/src/ontology/mondo-ingest.Makefile
@@ -213,3 +213,26 @@ deploy-mondo-ingest:
 	@test $(GHVERSION)
 	ls -alt $(DEPLOY_ASSETS_MONDO_INGEST)
 	gh release create $(GHVERSION) --notes "TBD." --title "$(GHVERSION)" --draft $(DEPLOY_ASSETS_MONDO_INGEST)
+
+tmp/mondo-ingest.owl:
+	wget https://github.com/monarch-initiative/mondo-ingest/releases/latest/download/mondo-ingest.owl -O $@
+
+tmp/mondo.owl:
+	wget http://purl.obolibrary.org/obo/mondo.owl -O $@
+
+tmp/mondo.sssom.tsv:
+	wget http://purl.obolibrary.org/obo/mondo/mappings/mondo.sssom.tsv -O $@
+
+tmp/mondo.sssom.ttl: tmp/mondo.sssom.tsv
+	sssom convert $< -O rdf -o $@
+
+# Merge Mondo, precise mappings and mondo-ingest into one coherent whole for the purpose of querying.
+
+tmp/merged.owl: tmp/mondo.owl tmp/mondo-ingest.owl tmp/mondo.sssom.ttl
+	$(ROBOT) merge -i tmp/mondo.owl -i tmp/mondo-ingest.owl -i tmp/mondo.sssom.ttl -o $@
+
+$(REPORTDIR)/mondo_ordo_unsupported_subclass.tsv: ../sparql/mondo-ordo-unsupported-subclass.sparql
+	$(ROBOT) query -i tmp/merged.owl --query $< $@
+
+.PHONY: mondo-ordo-subclass
+mondo-ordo-subclass: $(REPORTDIR)/mondo_ordo_unsupported_subclass.tsv
\ No newline at end of file
diff --git a/src/sparql/mondo-ordo-unsupported-subclass.sparql b/src/sparql/mondo-ordo-unsupported-subclass.sparql
new file mode 100644
index 000000000..aa188ffe5
--- /dev/null
+++ b/src/sparql/mondo-ordo-unsupported-subclass.sparql
@@ -0,0 +1,34 @@
+PREFIX skos: <http://www.w3.org/2004/02/skos/core#>
+PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+prefix IAO: <http://purl.obolibrary.org/obo/IAO_>
+prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+prefix oboInOwl: <http://www.geneontology.org/formats/oboInOwl#>
+prefix def: <http://purl.obolibrary.org/obo/IAO_0000115>
+prefix owl: <http://www.w3.org/2002/07/owl#>
+PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
+
+SELECT distinct ?sub_o1 ?sub_o2 ?super_o1 ?super_o2 ?sub #?source
+
+WHERE 
+{
+  	?sub_o1 rdfs:subClassOf* <http://purl.obolibrary.org/obo/MONDO_0000001> ;
+        rdfs:subClassOf ?super_o1 .
+    ?sub_o2 a owl:Class ;
+  		rdfs:subClassOf* <http://www.orpha.net/ORDO/Orphanet_377788> ;
+      rdfs:subClassOf ?super_o2 .
+  	?sub_o1 skos:exactMatch ?sub_o2 .
+  	?super_o1 skos:exactMatch ?super_o2 .
+   	?xref_anno a owl:Axiom ;
+            owl:annotatedSource ?sub_o1 ;
+            owl:annotatedProperty rdfs:subClassOf ;
+            owl:annotatedTarget ?super_o1 ;
+            oboInOwl:source ?source .
+  
+  FILTER(STRSTARTS(str(?source), "Orphanet:"))
+  
+  FILTER NOT EXISTS {
+   	?sub_o2 rdfs:subClassOf ?super_o2 .
+  }
+ #  FILTER(regex(str(?sub),"Orpha"))
+  FILTER( !isBlank(?sub_o1) && !isBlank(?sub_o2) && !isBlank(?super_o1) && !isBlank(?super_o2))
+} LIMIT 20