Update DrOn from RxNorm_full_01022024

src/ontology/catalog-v001.xml

@@ -29,6 +29,7 @@
 
   <!--  <uri name="http://purl.obolibrary.org/obo/dron/dron-upper.owl" uri="components/dron-upper.owl"/>-->
 
-
+  <uri name="http://purl.obolibrary.org/obo/dron/dron-obsolete.owl" uri="components/dron-obsolete.owl"/>
+
 </group>
-</catalog>
+</catalog>

src/ontology/dron-edit.owl

@@ -12,6 +12,7 @@ Ontology(<http://purl.obolibrary.org/obo/dron.owl>
 Import(<http://purl.obolibrary.org/obo/dron/dron-ingredient.owl>)
 Import(<http://purl.obolibrary.org/obo/dron/dron-ndc.owl>)
 Import(<http://purl.obolibrary.org/obo/dron/dron-rxnorm.owl>)
+Import(<http://purl.obolibrary.org/obo/dron/dron-obsolete.owl>)
 Import(<http://purl.obolibrary.org/obo/dron/imports/apollo_sv_import.owl>)
 Import(<http://purl.obolibrary.org/obo/dron/imports/bfo_import.owl>)
 Import(<http://purl.obolibrary.org/obo/dron/imports/chebi_import.owl>)
@@ -2160,4 +2161,4 @@ AnnotationAssertion(<http://purl.obolibrary.org/obo/IAO_0000412> <http://purl.ob
 
 
 AnnotationAssertion(<http://purl.obolibrary.org/obo/IAO_0000117> <http://purl.obolibrary.org/obo/DRON_00723856> "ciara lusnia")
-)
+)

src/ontology/dron.Makefile

@@ -11,7 +11,7 @@
 # In essence we merge rxnorm, dron-ingredient, all imports and the edit file
 # (no NDC) and the run a regular full release.
 
-LITE_ARTEFACTS=$(COMPONENTSDIR)/dron-rxnorm.owl $(COMPONENTSDIR)/dron-ingredient.owl $(IMPORT_OWL_FILES)
+LITE_ARTEFACTS=$(COMPONENTSDIR)/dron-rxnorm.owl $(COMPONENTSDIR)/dron-ingredient.owl $(COMPONENTSDIR)/dron-obsolete.owl $(IMPORT_OWL_FILES)
 $(TMPDIR)/dron-edit_lite.owl: $(SRC) $(LITE_ARTEFACTS)
 	$(ROBOT) remove --input $(SRC) --select imports \
 	merge $(patsubst %, -i %, $(LITE_ARTEFACTS)) --output [email protected] && mv [email protected] $@
@@ -33,13 +33,14 @@ $(TMPDIR)/ldtab.jar: | $(TMPDIR)
 LDTAB := java -jar $(TMPDIR)/ldtab.jar
 
 # Load DrOn templates into SQLite.
-$(TMPDIR)/dron.db: $(SCRIPTSDIR)/create-dron-tables.sql $(SCRIPTSDIR)/load-dron-tables.sql
+$(TMPDIR)/dron.db: $(SCRIPTSDIR)/create-dron-tables.sql $(SCRIPTSDIR)/load-dron-tables.sql $(SCRIPTSDIR)/index-dron-tables.sql
 	rm -f $@
 	sqlite3 $@ < $<
 	sqlite3 $@ < $(word 2,$^)
+	sqlite3 $@ < $(word 3,$^)
 
 # Load ChEBI into SQLite using LDTab.
-$(TMPDIR)/chebi.db: $(SCRIPTSDIR)/prefix.tsv $(TMPDIR)/mirror-chebi.owl | $(TMPDIR)/ldtab.jar
+$(TMPDIR)/chebi.db: $(SCRIPTSDIR)/prefix.tsv $(TMPDIR)/mirror-chebi.owl $(SCRIPTSDIR)/collect-chebi-labels.sql | $(TMPDIR)/ldtab.jar
 	$(eval DB=$@)
 	rm -f $@
 	$(LDTAB) init $(DB) --table chebi
@@ -48,6 +49,8 @@ $(TMPDIR)/chebi.db: $(SCRIPTSDIR)/prefix.tsv $(TMPDIR)/mirror-chebi.owl | $(TMPD
 	sqlite3 $(DB) "CREATE INDEX idx_chebi_subject ON chebi(subject)"
 	sqlite3 $(DB) "CREATE INDEX idx_chebi_predicate ON chebi(predicate)"
 	sqlite3 $(DB) "CREATE INDEX idx_chebi_object ON chebi(object)"
+	sqlite3 $(DB) "ANALYZE"
+	sqlite3 $(DB) < $(word 3,$^)
 
 # Create a SQLite database for RxNorm and load data from tmp/rxnorm/*.RRF.
 $(TMPDIR)/rxnorm.db: $(SCRIPTSDIR)/create-rxnorm-tables.sql $(SCRIPTSDIR)/load-rxnorm-tables.sql $(SCRIPTSDIR)/index-rxnorm-tables.sql | $(TMPDIR)/
@@ -61,13 +64,23 @@ update-labels: $(TMPDIR)/dron.db $(TMPDIR)/chebi.db $(TMPDIR)/rxnorm.db $(SCRIPT
 	sqlite3 < $(word 4,$^)
 	cd $(TEMPLATEDIR) && sqlite3 ../ontology/$< < ../ontology/$(word 5,$^)
 
+.PHONY: update-rxnorm
+update-rxnorm: $(TMPDIR)/dron.db $(TMPDIR)/chebi.db $(TMPDIR)/rxnorm.db $(SCRIPTSDIR)/update-dron-from-rxnorm.sql $(SCRIPTSDIR)/save-dron-tables.sql
+	sqlite3 < $(word 4,$^)
+	cd $(TEMPLATEDIR) && sqlite3 ../ontology/$< < ../ontology/$(word 5,$^)
+
+# Report common problems.
+$(TMPDIR)/problems.db: $(TMPDIR)/dron.db $(TMPDIR)/chebi.db $(TMPDIR)/rxnorm.db $(SCRIPTSDIR)/report-problems.sql
+	sqlite3 $@ < $(word 4,$^)
+
 # Convert DrOn template tables to LDTab format tables.
 $(TMPDIR)/ldtab.db: $(TMPDIR)/dron.db $(SCRIPTSDIR)/prefix.tsv $(SCRIPTSDIR)/convert-dron-ldtab.sql | $(TMPDIR)/ldtab.jar
 	$(eval DB=$@)
 	rm -f $@
 	$(LDTAB) init $(DB) --table dron_ingredient
 	$(LDTAB) init $(DB) --table dron_rxnorm
 	$(LDTAB) init $(DB) --table dron_ndc
+	$(LDTAB) init $(DB) --table dron_obsolete
 	$(LDTAB) prefix $@ $(word 2,$^)
 	sqlite3 $@ < $(word 3,$^)
 
@@ -81,7 +94,7 @@ $(COMPONENTSDIR)/%.owl: $(COMPONENTSDIR)/%.ttl
 	$(ROBOT) convert -i $< -o $@
 
 # Override the all_components task.
-all_components: $(COMPONENTSDIR)/dron-ingredient.ttl $(COMPONENTSDIR)/dron-rxnorm.ttl $(COMPONENTSDIR)/dron-ndc.ttl $(COMPONENTSDIR)/dron-ingredient.owl $(COMPONENTSDIR)/dron-rxnorm.owl $(COMPONENTSDIR)/dron-ndc.owl
+all_components: $(COMPONENTSDIR)/dron-ingredient.ttl $(COMPONENTSDIR)/dron-rxnorm.ttl $(COMPONENTSDIR)/dron-ndc.ttl $(COMPONENTSDIR)/dron-obsolete.ttl $(COMPONENTSDIR)/dron-ingredient.owl $(COMPONENTSDIR)/dron-rxnorm.owl $(COMPONENTSDIR)/dron-ndc.owl $(COMPONENTSDIR)/dron-obsolete.owl
 
 ###################################
 #### Create templates from OWL ####
@@ -99,17 +112,19 @@ $(TMPDIR)/reverse/:
 $(TMPDIR)/reverse/dron-%.owl: components.bk/dron-%.owl | $(TMPDIR)/reverse/
 	$(ROBOT) convert -i $< -o $@
 
-$(TMPDIR)/reverse.db: $(SCRIPTSDIR)/prefix.tsv $(SCRIPTSDIR)/create-dron-tables.sql $(SCRIPTSDIR)/convert-ldtab-dron.sql $(TMPDIR)/reverse/dron-ingredient.owl $(TMPDIR)/reverse/dron-rxnorm.owl $(TMPDIR)/reverse/dron-ndc.owl | $(TMPDIR)/ldtab.jar
+$(TMPDIR)/reverse.db: $(SCRIPTSDIR)/prefix.tsv $(SCRIPTSDIR)/create-dron-tables.sql $(SCRIPTSDIR)/convert-ldtab-dron.sql $(TMPDIR)/reverse/dron-ingredient.owl $(TMPDIR)/reverse/dron-rxnorm.owl $(TMPDIR)/reverse/dron-ndc.owl $(TMPDIR)/reverse/dron-obsolete.owl | $(TMPDIR)/ldtab.jar
 	$(eval DB=$@)
 	rm -f $(DB)
 	$(LDTAB) init $(DB) --table dron_ingredient
 	$(LDTAB) init $(DB) --table dron_rxnorm
 	$(LDTAB) init $(DB) --table dron_ndc
+	$(LDTAB) init $(DB) --table dron_obsolete
 	$(LDTAB) prefix $(DB) $<
 	sqlite3 $(DB) < $(word 2,$^)
 	$(LDTAB) import $(DB) --table dron_ingredient $(TMPDIR)/reverse/dron-ingredient.owl
 	$(LDTAB) import $(DB) --table dron_rxnorm $(TMPDIR)/reverse/dron-rxnorm.owl
 	$(LDTAB) import $(DB) --table dron_ndc $(TMPDIR)/reverse/dron-ndc.owl
+	$(LDTAB) import $(DB) --table dron_obsolete $(TMPDIR)/reverse/dron-obsolete.owl
 
 .PHONY: reverse
 reverse: $(TMPDIR)/reverse.db $(SCRIPTSDIR)/save-dron-tables.sql
@@ -143,4 +158,4 @@ $(TMPDIR)/%.owl.diff: $(TMPDIR)/reverse/%.owl $(COMPONENTSDIR)/%.owl
 	-diff -u $^ > $@
 
 .PHONY: roundtrip
-roundtrip: $(COMPONENTSDIR)/dron-ingredient.ttl $(COMPONENTSDIR)/dron-rxnorm.ttl $(COMPONENTSDIR)/dron-ndc.ttl $(TMPDIR)/dron-ingredient.owl.diff $(TMPDIR)/dron-ingredient.tsv.diff $(TMPDIR)/dron-rxnorm.owl.diff $(TMPDIR)/dron-rxnorm.tsv.diff $(TMPDIR)/dron-ndc.owl.diff $(TMPDIR)/dron-ndc.tsv.diff
+roundtrip: $(COMPONENTSDIR)/dron-ingredient.ttl $(COMPONENTSDIR)/dron-rxnorm.ttl $(COMPONENTSDIR)/dron-ndc.ttl $(COMPONENTSDIR)/dron-obsolete.ttl $(TMPDIR)/dron-ingredient.owl.diff $(TMPDIR)/dron-ingredient.tsv.diff $(TMPDIR)/dron-rxnorm.owl.diff $(TMPDIR)/dron-rxnorm.tsv.diff $(TMPDIR)/dron-ndc.owl.diff $(TMPDIR)/dron-ndc.tsv.diff $(TMPDIR)/dron-obsolete.owl.diff $(TMPDIR)/dron-obsolete.tsv.diff

src/scripts/collect-chebi-labels.sql

@@ -0,0 +1,20 @@
+-- Collect labels from ChEBI terms that are not deprecated.
+DROP TABLE IF EXISTS label;
+CREATE TABLE label (
+    curie TEXT PRIMARY KEY,
+    label TEXT UNIQUE
+);
+
+INSERT OR IGNORE INTO label
+SELECT
+  subject AS curie,
+  object AS label
+FROM chebi
+WHERE predicate = 'rdfs:label'
+  AND subject NOT IN (
+    SELECT subject
+    FROM chebi
+    WHERE predicate = 'owl:deprecated'
+  );
+CREATE INDEX idx_label_label ON label(label);
+ANALYZE label;

src/scripts/convert-dron-ldtab.sql

@@ -474,3 +474,38 @@ FROM (
     UNION
     SELECT curie, clinical_drug AS drug FROM dron.ndc_clinical_drug
 );
+
+
+-- ## Obsolete Terms
+
+INSERT OR IGNORE INTO dron_obsolete(subject, predicate, object)
+SELECT
+    curie AS subject,
+    'rdf:type' AS predicate,
+    type AS object
+FROM obsolete;
+
+INSERT OR IGNORE INTO dron_obsolete(subject, predicate, object, datatype)
+SELECT
+    curie AS subject,
+    'rdfs:label' AS predicate,
+    label AS object,
+    'xsd:string' AS datatype
+FROM obsolete;
+
+INSERT OR IGNORE INTO dron_obsolete(subject, predicate, object, datatype)
+SELECT
+    curie AS subject,
+    'owl:deprecated' AS predicate,
+    'true' AS object,
+    'xsd:boolean' AS datatype
+FROM obsolete;
+
+INSERT OR IGNORE INTO dron_obsolete(subject, predicate, object)
+SELECT
+    curie AS subject,
+    'IAO:0100001' AS predicate,
+    replaced_by AS object
+FROM obsolete
+WHERE replaced_by IS NOT NULL
+  AND replaced_by != '';

src/scripts/convert-ldtab-dron.sql

@@ -174,3 +174,22 @@ UNION
 SELECT rxcui FROM clinical_drug
 UNION
 SELECT rxcui FROM branded_drug;
+
+
+-- ## dron-obsolete
+
+INSERT OR IGNORE INTO obsolete(curie, label, type)
+SELECT
+    a.subject AS curie,
+    b.object AS label,
+    a.object AS type
+FROM dron_obsolete AS a
+LEFT JOIN dron_obsolete AS b ON a.subject = b.subject
+WHERE a.predicate = 'rdf:type'
+  AND b.predicate = 'rdfs:label';
+
+UPDATE obsolete
+SET replaced_by = object
+FROM dron_obsolete
+WHERE obsolete.curie = dron_obsolete.subject
+  AND dron_obsolete.predicate = 'IAO:0100001';

src/scripts/create-dron-tables.sql

@@ -1,104 +1,116 @@
 -- Create the DrOn tables.
 
 CREATE TABLE rxcui (
-    rxcui INT PRIMARY KEY,
+    rxcui INTEGER PRIMARY KEY NOT NULL,
     replaced_by INTEGER
 );
 
 CREATE TABLE disposition (
     curie TEXT PRIMARY KEY,
-    label TEXT UNIQUE
+    label TEXT UNIQUE NOT NULL
 );
 
 CREATE TABLE ingredient (
     curie TEXT PRIMARY KEY,
-    label TEXT, -- UNIQUE,
-    rxcui INTEGER,
+    label TEXT NOT NULL,
+    rxcui INTEGER NOT NULL,
     FOREIGN KEY (rxcui) REFERENCES rxcui(rxcui)
 );
 
 CREATE TABLE ingredient_disposition (
-    ingredient TEXT,
-    disposition TEXT,
+    ingredient TEXT NOT NULL,
+    disposition TEXT NOT NULL,
     FOREIGN KEY (ingredient) REFERENCES ingredient(curie),
     FOREIGN KEY (disposition) REFERENCES disposition(curie)
 );
 
 CREATE TABLE clinical_drug_form (
     curie TEXT PRIMARY KEY,
-    label TEXT, -- UNIQUE,
+    label TEXT NOT NULL, -- UNIQUE,
     parent TEXT,
-    rxcui INTEGER,
+    rxcui INTEGER NOT NULL,
     FOREIGN KEY (rxcui) REFERENCES rxcui(rxcui)
 );
 
 CREATE TABLE clinical_drug_form_ingredient (
-    clinical_drug_form TEXT,
-    ingredient TEXT,
+    clinical_drug_form TEXT NOT NULL,
+    ingredient TEXT NOT NULL,
+    PRIMARY KEY (clinical_drug_form, ingredient),
     FOREIGN KEY (clinical_drug_form) REFERENCES clinical_drug_form(curie),
     FOREIGN KEY (ingredient) REFERENCES ingredient(curie)
 );
 
 CREATE TABLE clinical_drug_form_disposition (
-    clinical_drug_form TEXT,
-    disposition TEXT,
+    clinical_drug_form TEXT NOT NULL,
+    disposition TEXT NOT NULL,
+    PRIMARY KEY (clinical_drug_form, disposition),
     FOREIGN KEY (clinical_drug_form) REFERENCES clinical_drug_form(curie),
     FOREIGN KEY (disposition) REFERENCES disposition(curie)
 );
 
 CREATE TABLE clinical_drug (
     curie TEXT PRIMARY KEY,
-    label TEXT, -- UNIQUE,
-    clinical_drug_form TEXT,
-    rxcui INTEGER,
+    label TEXT NOT NULL, -- UNIQUE,
+    clinical_drug_form TEXT NOT NULL,
+    rxcui INTEGER NOT NULL,
     FOREIGN KEY (clinical_drug_form) REFERENCES clinical_drug_form(curie),
     FOREIGN KEY (rxcui) REFERENCES rxcui(rxcui)
 );
 
 CREATE TABLE clinical_drug_strength (
-    clinical_drug TEXT,
-    ingredient TEXT,
-    strength TEXT,
-    unit TEXT,
+    clinical_drug TEXT NOT NULL,
+    ingredient TEXT NOT NULL,
+    strength TEXT NOT NULL,
+    unit TEXT NOT NULL,
+    -- WARN: There are some clinical drugs in RxNorm
+    -- with the same ingredient but multiple strengths.
+    PRIMARY KEY (clinical_drug, ingredient, strength, unit),
     FOREIGN KEY (clinical_drug) REFERENCES clinical_drug(curie),
     FOREIGN KEY (ingredient) REFERENCES ingredient(curie)
 );
 
 CREATE TABLE branded_drug (
     curie TEXT PRIMARY KEY,
-    label TEXT, -- UNIQUE,
-    clinical_drug TEXT,
-    rxcui INTEGER,
+    label TEXT NOT NULL, -- UNIQUE,
+    clinical_drug TEXT NOT NULL,
+    rxcui INTEGER NOT NULL,
     FOREIGN KEY (clinical_drug) REFERENCES clinical_drug(curie),
     FOREIGN KEY (rxcui) REFERENCES rxcui(rxcui)
 );
 
 CREATE TABLE branded_drug_excipient (
-    branded_drug TEXT,
-    ingredient TEXT,
+    branded_drug TEXT NOT NULL,
+    ingredient TEXT NOT NULL,
+    PRIMARY KEY (branded_drug, ingredient),
     FOREIGN KEY (branded_drug) REFERENCES branded_drug(curie),
     FOREIGN KEY (ingredient) REFERENCES ingredient(curie)
 );
 
 CREATE TABLE ndc_branded_drug (
     curie TEXT PRIMARY KEY,
-    ndc TEXT,
-    branded_drug TEXT,
+    ndc TEXT UNIQUE NOT NULL,
+    branded_drug TEXT NOT NULL,
     FOREIGN KEY (branded_drug) REFERENCES branded_drug(curie)
 );
 
 CREATE TABLE ndc_clinical_drug (
     curie TEXT PRIMARY KEY,
-    ndc TEXT,
-    clinical_drug TEXT,
+    ndc TEXT UNIQUE NOT NULL,
+    clinical_drug TEXT NOT NULL,
     FOREIGN KEY (clinical_drug) REFERENCES clinical_drug(curie)
 );
 
+CREATE TABLE obsolete (
+    curie TEXT PRIMARY KEY NOT NULL,
+    label TEXT NOT NULL, -- UNIQUE,
+    type TEXT NOT NULL,
+    replaced_by TEXT
+);
 
 -- Create triggers for automatically assigning DRON IDs.
 -- First create a table with one row to track the current ID.
 CREATE TABLE current_dron_id ( id INTEGER );
-INSERT INTO current_dron_id VALUES (1000000); -- TODO: check this!
+-- INSERT INTO current_dron_id VALUES (1003000);
 
 -- Then create a trigger for each row where a CURIE can be assigned:
 -- When a row is inserted into the table,

src/scripts/index-dron-tables.sql

@@ -0,0 +1,36 @@
+-- Index main DrOn tables.
+
+CREATE INDEX IF NOT EXISTS idx_ingredient_curie ON ingredient(curie);
+CREATE INDEX IF NOT EXISTS idx_ingredient_label ON ingredient(label);
+CREATE INDEX IF NOT EXISTS idx_ingredient_rxcui ON ingredient(rxcui);
+
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_form_curie ON clinical_drug_form(curie);
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_form_label ON clinical_drug_form(label);
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_form_rxcui ON clinical_drug_form(rxcui);
+
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_form_ingredient_clinical_drug_form ON clinical_drug_form_ingredient(clinical_drug_form);
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_form_ingredient_ingredient ON clinical_drug_form_ingredient(ingredient);
+
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_curie ON clinical_drug(curie);
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_label ON clinical_drug(label);
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_rxcui ON clinical_drug(rxcui);
+
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_strength_clinical_drug ON clinical_drug_strength(clinical_drug);
+CREATE INDEX IF NOT EXISTS idx_clinical_drug_strength_ingredient ON clinical_drug_strength(ingredient);
+
+CREATE INDEX IF NOT EXISTS idx_branded_drug_curie ON branded_drug(curie);
+CREATE INDEX IF NOT EXISTS idx_branded_drug_label ON branded_drug(label);
+CREATE INDEX IF NOT EXISTS idx_branded_drug_rxcui ON branded_drug(rxcui);
+
+CREATE INDEX IF NOT EXISTS idx_branded_drug_excipient_branded_drug ON branded_drug_excipient(branded_drug);
+CREATE INDEX IF NOT EXISTS idx_branded_drug_excipient_ingredient ON branded_drug_excipient(ingredient);
+
+CREATE INDEX IF NOT EXISTS idx_ndc_clinical_drug_curie ON ndc_clinical_drug(curie);
+CREATE INDEX IF NOT EXISTS idx_ndc_clinical_drug_ndc ON ndc_clinical_drug(ndc);
+CREATE INDEX IF NOT EXISTS idx_ndc_clinical_drug_clinical_drug ON ndc_clinical_drug(clinical_drug);
+
+CREATE INDEX IF NOT EXISTS idx_ndc_branded_drug_curie ON ndc_branded_drug(curie);
+CREATE INDEX IF NOT EXISTS idx_ndc_branded_drug_ndc ON ndc_branded_drug(ndc);
+CREATE INDEX IF NOT EXISTS idx_ndc_branded_drug_branded_drug ON ndc_branded_drug(branded_drug);
+
+ANALYZE;