Improve automated Zooma mappings

cmat/trait_mapping/main.py

@@ -47,7 +47,7 @@ def process_trait(trait: Trait, filters: dict, zooma_host: str, oxo_target_list:
     """
     logger.debug('Processing trait {}'.format(trait.name))
 
-    trait.zooma_result_list = get_zooma_results(trait.name, filters, zooma_host, target_ontology)
+    trait.zooma_result_list = get_zooma_results(trait.name.lower(), filters, zooma_host, target_ontology)
     trait.process_zooma_results()
     if (trait.is_finished
             or len(trait.zooma_result_list) == 0

cmat/trait_mapping/trait.py

@@ -58,12 +58,11 @@ def process_zooma_results(self):
         Check whether any Zooma mappings can be output as a finished ontology mapping.
         Put any finished mappings in finished_mapping_set
         """
-        for mapping in self.zooma_result_list:
-            if mapping.confidence.lower() != "high":
-                continue
-
-            for mapping in mapping.mapping_list:
-                if mapping.in_ontology and mapping.is_current:
+        for zooma_result in self.zooma_result_list:
+            for mapping in zooma_result.mapping_list:
+                # Accept either high-confidence mappings, or exact string matches
+                if mapping.in_ontology and mapping.is_current and (zooma_result.confidence.lower() == "high"
+                                                                   or zooma_result.zooma_label.lower() == self.name.lower()):
                     ontology_entry = OntologyEntry(mapping.uri, mapping.ontology_label)
                     self.finished_mapping_set.add(ontology_entry)
 

tests/pipelines/resources/expected/consequences_snp.tsv

@@ -411,6 +411,7 @@
 17:65558594:G:A	ENSG00000168646	AXIN2	synonymous_variant
 17:6556374:C:T	ENSG00000091622	PITPNM3	intron_variant
 17:6628328:A:G	ENSG00000198920	KIAA0753	synonymous_variant
+17:6628328:A:G	ENSG00000282936		3_prime_UTR_variant
 17:6693178:A:AACACACACACAC	ENSG00000141485	SLC13A5	splice_polypyrimidine_tract_variant
 17:7224245:T:C	ENSG00000072778	ACADVL	splice_donor_variant
 17:73201307:A:G	ENSG00000166685	COG1	missense_variant
@@ -900,7 +901,7 @@
 4:120785105:A:G	ENSG00000138738	PRDM5	splice_polypyrimidine_tract_variant
 4:121847458:G:T	ENSG00000138686	BBS7	missense_variant
 4:121853086:C:T	ENSG00000138686	BBS7	missense_variant
-4:122979387:G:T	ENSG00000145375	SPATA5	splice_donor_variant
+4:122979387:G:T	ENSG00000145375	AFG2A	splice_donor_variant
 4:127881908:G:T	ENSG00000142731	PLK4	missense_variant
 4:127930737:G:T	ENSG00000164073	MFSD8	missense_variant
 4:128959797:C:A	ENSG00000151466	SCLT1	intron_variant

tests/pipelines/resources/expected/evidence_strings.json

@@ -788,6 +788,7 @@
 {"alleleOrigins": ["germline"], "datasourceId": "eva", "datatypeId": "genetic_association", "clinicalSignificances": ["uncertain significance"], "confidence": "criteria provided, single submitter", "studyId": "RCV001373139", "releaseDate": "2021-04-13", "targetFromSourceId": "ENSG00000163930", "variantFunctionalConsequenceId": "SO_0001583", "variantId": "3_52402628_G_A", "cohortPhenotypes": ["BAP1 tumor predisposition syndrome", "BAP1-related tumor predisposition syndrome", "Tumor predisposition syndrome", "Tumor susceptibility linked to germline BAP1 mutations"], "diseaseFromSource": "BAP1-related tumor predisposition syndrome", "diseaseFromSourceId": "C3280492", "diseaseFromSourceMappedId": "MONDO_0013692", "variantHgvsId": "NC_000003.12:g.52402628G>A"}
 {"alleleOrigins": ["germline"], "datasourceId": "eva", "datatypeId": "genetic_association", "clinicalSignificances": ["uncertain significance"], "confidence": "criteria provided, single submitter", "studyId": "RCV001373139", "releaseDate": "2021-04-13", "targetFromSourceId": "ENSG00000163930", "variantFunctionalConsequenceId": "SO_0001583", "variantId": "3_52402628_G_A", "cohortPhenotypes": ["BAP1 tumor predisposition syndrome", "BAP1-related tumor predisposition syndrome", "Tumor predisposition syndrome", "Tumor susceptibility linked to germline BAP1 mutations"], "diseaseFromSource": "BAP1-related tumor predisposition syndrome", "diseaseFromSourceId": "C3280492", "diseaseFromSourceMappedId": "Orphanet_289539", "variantHgvsId": "NC_000003.12:g.52402628G>A"}
 {"alleleOrigins": ["germline"], "datasourceId": "eva", "datatypeId": "genetic_association", "clinicalSignificances": ["benign"], "confidence": "criteria provided, single submitter", "studyId": "RCV001730858", "releaseDate": "2022-02-20", "targetFromSourceId": "ENSG00000198920", "variantFunctionalConsequenceId": "SO_0001819", "variantId": "17_6628328_A_G", "cohortPhenotypes": ["Joubert syndrome 38"], "diseaseFromSource": "Joubert syndrome 38", "diseaseFromSourceId": "C5561958", "variantHgvsId": "NC_000017.11:g.6628328A>G"}
+{"alleleOrigins": ["germline"], "datasourceId": "eva", "datatypeId": "genetic_association", "clinicalSignificances": ["benign"], "confidence": "criteria provided, single submitter", "studyId": "RCV001730858", "releaseDate": "2022-02-20", "targetFromSourceId": "ENSG00000282936", "variantFunctionalConsequenceId": "SO_0001624", "variantId": "17_6628328_A_G", "cohortPhenotypes": ["Joubert syndrome 38"], "diseaseFromSource": "Joubert syndrome 38", "diseaseFromSourceId": "C5561958", "variantHgvsId": "NC_000017.11:g.6628328A>G"}
 {"alleleOrigins": ["germline"], "datasourceId": "eva", "datatypeId": "genetic_association", "clinicalSignificances": ["likely benign"], "confidence": "criteria provided, single submitter", "studyId": "RCV002057147", "releaseDate": "2022-06-09", "targetFromSourceId": "ENSG00000115904", "variantFunctionalConsequenceId": "SO_0001627", "variantId": "2_39120324_C_A", "variantRsId": "rs368569135", "cohortPhenotypes": ["Noonan spectrum disorder", "RASopathy", "rasopathies"], "diseaseFromSource": "RASopathy", "diseaseFromSourceId": "C5555857", "diseaseFromSourceMappedId": "EFO_1001502", "variantHgvsId": "NC_000002.12:g.39120324C>A"}
 {"alleleOrigins": ["germline"], "datasourceId": "eva", "datatypeId": "genetic_association", "clinicalSignificances": ["pathogenic"], "confidence": "criteria provided, single submitter", "studyId": "RCV002247286", "releaseDate": "2022-06-09", "targetFromSourceId": "ENSG00000198712", "variantFunctionalConsequenceId": "SO_0001631", "variantId": "MT_7512_T_C", "variantRsId": "rs199474817", "cohortPhenotypes": ["COX deficiency", "Complex 4 mitochondrial respiratory chain deficiency", "Complex IV deficiency", "Cytochrome-c oxidase deficiency", "Cytochrome-c oxidase deficiency disease", "Deficiency of mitochondrial respiratory chain complex4", "MITOCHONDRIAL COMPLEX IV DEFICIENCY, NUCLEAR TYPE 1", "Mitochondrial complex IV deficiency"], "diseaseFromSource": "Cytochrome-c oxidase deficiency disease", "diseaseFromSourceId": "C5435656", "diseaseFromSourceMappedId": "MONDO_0009068", "variantHgvsId": "NC_012920.1:m.7512T>C"}
 {"alleleOrigins": ["germline"], "datasourceId": "eva", "datatypeId": "genetic_association", "clinicalSignificances": ["pathogenic"], "confidence": "criteria provided, single submitter", "studyId": "RCV002247286", "releaseDate": "2022-06-09", "targetFromSourceId": "ENSG00000198786", "variantFunctionalConsequenceId": "SO_0001631", "variantId": "MT_7512_T_C", "variantRsId": "rs199474817", "cohortPhenotypes": ["COX deficiency", "Complex 4 mitochondrial respiratory chain deficiency", "Complex IV deficiency", "Cytochrome-c oxidase deficiency", "Cytochrome-c oxidase deficiency disease", "Deficiency of mitochondrial respiratory chain complex4", "MITOCHONDRIAL COMPLEX IV DEFICIENCY, NUCLEAR TYPE 1", "Mitochondrial complex IV deficiency"], "diseaseFromSource": "Cytochrome-c oxidase deficiency disease", "diseaseFromSourceId": "C5435656", "diseaseFromSourceMappedId": "MONDO_0009068", "variantHgvsId": "NC_012920.1:m.7512T>C"}

tests/trait_mapping/test_main.py

@@ -6,7 +6,8 @@
 
 import pytest
 
-from cmat.trait_mapping.main import parse_traits, process_traits
+from cmat.trait_mapping.main import parse_traits, process_traits, process_trait
+from cmat.trait_mapping.trait import Trait
 
 
 def get_test_resource(resource_name):
@@ -62,3 +63,27 @@ def test_main():
     mapped_terms = {x[0] for x in output_mappings}
     curation_terms = {x[0] for x in output_curation}
     assert len(mapped_terms) + len(curation_terms) == len(all_terms)
+
+
+def test_process_trait_exact_match():
+    # Exact match with MONDO:0009061 (in EFO and Mondo)
+    trait_name = 'Cystic Fibrosis'
+    # Don't use any data sources in Zooma as those will come back as high-confidence matches
+    zooma_filters = {'ontologies': 'efo,mondo,hp',
+                     'required': 'none',
+                     'preferred': 'none'}
+    zooma_host = 'https://www.ebi.ac.uk'
+    # Don't use OxO
+    oxo_targets = []
+    oxo_distance = 0
+
+    # This should be marked as finished, as it's an exact string match with a term contained in the target ontology
+    efo_trait = process_trait(Trait(trait_name, None, None), zooma_filters, zooma_host, oxo_targets, oxo_distance,
+                              target_ontology='efo')
+    assert efo_trait.is_finished
+
+    # This should not be marked as finished, even though Zooma finds an exact match in one of its ontologies, it's not
+    # the requested target ontology and thus still needs to be curated
+    hpo_trait = process_trait(Trait(trait_name, None, None), zooma_filters, zooma_host, oxo_targets, oxo_distance,
+                              target_ontology='hp')
+    assert not hpo_trait.is_finished