add mutations and molecular profiles

cbioportal/adapters/cbioportal_adapter.py

@@ -62,7 +62,6 @@ def get_data_from_api(self, _type):
         if _type in [SampleField, samplePatientAssociationField]:
             if "samples" in self.api_called: return self.api_called["samples"]
             self.api_called["samples"] = []
-            # studies = self.cbioportal.Studies.getAllStudiesUsingGET().result()
             studies = self.check_api_called(StudyField)
             print(f"Getting samples of {len(studies)} studies")
             for i, study in enumerate(studies):
@@ -75,12 +74,11 @@ def get_data_from_api(self, _type):
             if "sample_lists" in self.api_called: return self.api_called["sample_lists"]
             self.api_called["sample_lists"] = self.cbioportal.Sample_Lists.getAllSampleListsUsingGET().result()[0:self.limit]
             return self.api_called["sample_lists"]
-
         if _type == GeneField:
             if "genes" in self.api_called: return self.api_called["genes"]
             self.api_called["genes"] = self.cbioportal.Genes.getAllGenesUsingGET().result()[0:self.limit]
             return self.api_called["genes"]
-        if _type in [GenePanelField, GenePanelGeneAssociationField]:
+        if _type in [GenePanelField]:
             if "gene_panels" in self.api_called: return self.api_called["gene_panels"]
             self.api_called["gene_panels"] = self.cbioportal.Gene_Panels.getAllGenePanelsUsingGET().result()[0:self.limit]
             return self.api_called["gene_panels"]
@@ -92,36 +90,42 @@ def get_data_from_api(self, _type):
             if "clinical_attributes" in self.api_called: return self.api_called["clinical_attributes"]
             self.api_called["clinical_attributes"] = self.cbioportal.Clinical_Attributes.getAllClinicalAttributesUsingGET().result()[0:self.limit]
             return self.api_called["clinical_attributes"]
-        if _type in [PatientSampleStudyEntityField, PatientToPatientSampleStudyEntityField, SampleToPatientSampleStudyEntityField, StudyToPatientSampleStudyEntityField]:
-            print("not correctly implemeted yet!")
-            return []
 
-            if "patient_sample_study_entities" in self.api_called: return self.api_called["patient_sample_study_entities"]
-            patients = self.check_api_called(PatientField)
-            samples = self.check_api_called(SampleField)
-            studies = self.check_api_called(StudyField)
-            self.api_called["patient_sample_study_entities"] = []
-            counter = 0
-            for _id1 in patients:
-                for _id2 in samples:
-                    for _id3 in studies:
-                        id = hash(f"{_id1.patientId}_{_id2.sampleId}_{_id3.studyId}")
-                        self.api_called["patient_sample_study_entities"].append({"patientId":_id1.patientId, "sampleId":_id2.sampleId, "studyId":_id3.studyId, "id":id})
-                        counter += 1
-                        if counter == self.limit:
-                            return self.api_called["patient_sample_study_entities"]
-            return self.api_called["patient_sample_study_entities"]
-
+        if _type in [MutationField, mutationToSampleField, mutationToGeneField, mutationToStudyField, mutationToPatientField, mutationToMolecularProfileField]:
+            if "mutations" in self.api_called: return self.api_called["mutations"]
+            profiles = self.check_api_called(MolecularProfileField)
+            # get a list of all molecular profile ids of all the profiles
+            molecularProfileIds = [profile["molecularProfileId"] for profile in profiles]
+            print(f"Getting mutations in {len(molecularProfileIds)} molecular profiles")
+            mutations_per_profile = {}
+            def get_mutations_in_profile(profile_id):
+                try:
+                    return self.cbioportal.Mutations.fetchMutationsInMultipleMolecularProfilesUsingPOST(mutationMultipleStudyFilter = {"molecularProfileIds":[profile_id]}).result()[0:self.limit]
+                except HTTPNotFound:
+                    return []
+            for i, profile_id in enumerate(molecularProfileIds):
+                mutations_per_profile[profile_id] = get_mutations_in_profile(profile_id)
+                print(f"Profile {i+1}/{len(molecularProfileIds)}: {len(mutations_per_profile[profile_id])} mutations")
+
+            self.api_called["mutations"] = [mutation for mutations in mutations_per_profile.values() for mutation in mutations]
+            print(f"found {len(self.api_called['mutations'])} mutations in {len(molecularProfileIds)} molecular profiles")
+            return self.api_called["mutations"]
         if _type == "test":
             print(dir(self.cbioportal))
-            gene_panels = self.cbioportal.Gene_Panels.getAllGenePanelsUsingGET().result()
-            print(gene_panels[0])
+
+
         else:
             raise ValueError(f"Node type {_type} not supported.")
 
     def _yield_node_type(self, items, node_type):
         for item in items:
-            _id = item[node_type._ID.value]
+            if isinstance(node_type._ID.value, str):
+                _id = item[node_type._ID.value] 
+            elif isinstance(node_type._ID.value, list):
+                _id = str(hash("_".join([str(item[field]) for field in node_type._ID.value])))
+            else:
+                raise ValueError("ID field must be a string or a list of strings.")
+
             _type = node_type._LABEL.value
             _props = {"version": self.version}
             for field in node_type:
@@ -151,8 +155,19 @@ def get_nodes(self):
 
     def _yield_edge_type(self, items, edge_type):
         for item in items:
-            _subject = item[edge_type._SUBJECT.value]
-            _object = item[edge_type._OBJECT.value]
+            if isinstance(edge_type._SUBJECT.value, str):
+                _subject = item[edge_type._SUBJECT.value] 
+            elif isinstance(edge_type._SUBJECT.value, list):
+                _subject = str(hash("_".join([str(item[field]) for field in edge_type._SUBJECT.value])))
+            else:
+                raise ValueError("_subject field must be a string or a list of strings.")
+
+            if isinstance(edge_type._OBJECT.value, str):
+                _object = item[edge_type._OBJECT.value] 
+            elif isinstance(edge_type._OBJECT.value, list):
+                _object = str(hash("_".join([str(item[field]) for field in edge_type._OBJECT.value])))
+            else:
+                raise ValueError("_subject field must be a string or a list of strings.")
 
             try:
                 item[edge_type._ID.value]

cbioportal/adapters/edge_field_classes.py

@@ -35,14 +35,6 @@ class samplePatientAssociationField(Enum):
     _OBJECT = "patientId"
     _LABEL = "fromPatient"
 
-class GenePanelGeneAssociationField(Enum):
-    """
-    Define possible fields the adapter can provide for genePanel-gene associations.
-    """
-    _SUBJECT = "genePanelId"
-    _OBJECT = "genes"
-    _LABEL = "hasGene"
-
 class MolecularProfiletoStudyField(Enum):
     """
     Define possible fields the adapter can provide for molecular profile - study associations.
@@ -75,27 +67,42 @@ class CopyNumberSegmentToSampleField(Enum):
     _OBJECT = "sampleId"
     _LABEL = "fromSample"
 
+class mutationToSampleField(Enum):
+    """
+    Define possible fields the adapter can provide for mutation - sample associations.
+    """
+    _SUBJECT = ["molecularProfileId", "sampleId", "patientId", "entrezGeneId", "studyId"]
+    _OBJECT = "sampleId"
+    _LABEL = "fromSample"
 
-class PatientToPatientSampleStudyEntityField(Enum):
+class mutationToGeneField(Enum):
     """
-    Define possible fields the adapter can provide for patient - patient sample study entity associations.
+    Define possible fields the adapter can provide for mutation - gene associations.
     """
-    _SUBJECT = "patientId"
-    _OBJECT = "id"
-    _LABEL = "partOf"
+    _SUBJECT = ["molecularProfileId", "sampleId", "patientId", "entrezGeneId", "studyId"]
+    _OBJECT = "entrezGeneId"
+    _LABEL = "fromGene"
 
-class SampleToPatientSampleStudyEntityField(Enum):
+class mutationToStudyField(Enum):
     """
-    Define possible fields the adapter can provide for sample - patient sample study entity associations.
+    Define possible fields the adapter can provide for mutation - study associations.
     """
-    _SUBJECT = "sampleId"
-    _OBJECT = "id"
-    _LABEL = "partOf"
+    _SUBJECT = ["molecularProfileId", "sampleId", "patientId", "entrezGeneId", "studyId"]
+    _OBJECT = "studyId"
+    _LABEL = "hasStudy"
 
-class StudyToPatientSampleStudyEntityField(Enum):
+class mutationToPatientField(Enum):
     """
-    Define possible fields the adapter can provide for study - patient sample study entity associations.
+    Define possible fields the adapter can provide for mutation - patient associations.
     """
-    _SUBJECT = "studyId"
-    _OBJECT = "id"
-    _LABEL = "partOf"
+    _SUBJECT = ["molecularProfileId", "sampleId", "patientId", "entrezGeneId", "studyId"]
+    _OBJECT = "patientId"
+    _LABEL = "fromPatient"
+
+class mutationToMolecularProfileField(Enum):
+    """
+    Define possible fields the adapter can provide for mutation - molecular profile associations.
+    """
+    _SUBJECT = ["molecularProfileId", "sampleId", "patientId", "entrezGeneId", "studyId"]
+    _OBJECT = "molecularProfileId"
+    _LABEL = "fromMolecularProfile"

cbioportal/adapters/node_field_classes.py

@@ -48,7 +48,6 @@ class SampleListField(Enum):
     NAME = "name"
     CATEGORY = "category"
     SAMPLE_COUNT = "sampleCount"
-    SAMPLE_IDS = "sampleIds"
 
 class GeneField(Enum):
     _ID = "entrezGeneId"
@@ -111,14 +110,38 @@ class CopyNumberSegmentField(Enum):
     SEGMENT_MEAN = "segmentMean"
     START = "start"
     STUDY_ID = "studyId"
-
 
-class PatientSampleStudyEntityField(Enum):
-    """
-    Define possible fields the adapter can provide for patient-sample-study entities.
-    """
-    PATIENT_ID = "patientId"
-    SAMPLE_ID = "sampleId"
-    STUDY_ID = "studyId"
-    _LABEL = "PatientSampleStudyEntity"
-    _ID = "id"
+# EXAMPLE
+# Mutation(alleleSpecificCopyNumber=None, aminoAcidChange=None, center='MSKCC', chr='15', driverFilter=None, driverFilterAnnotation=None, driverTiersFilter=None, driverTiersFilterAnnotation=None, endPosition=30700159, entrezGeneId=101059918, gene=None, keyword='GOLGA8R W275 missense', molecularProfileId='acbc_mskcc_2015_mutations', mutationStatus='SOMATIC', mutationType='Missense_Mutation', namespaceColumns=None, ncbiBuild='GRCh37', normalAltCount=None, normalRefCount=None, patientId='AdCC3T', proteinChange='W275R', proteinPosEnd=275, proteinPosStart=275, referenceAllele='A', refseqMrnaId='NM_001282484.1', sampleId='AdCC3T', startPosition=30700159, studyId='acbc_mskcc_2015', tumorAltCount=20, tumorRefCount=80, uniquePatientKey='QWRDQzNUOmFjYmNfbXNrY2NfMjAxNQ', uniqueSampleKey='QWRDQzNUOmFjYmNfbXNrY2NfMjAxNQ', validationStatus='Unknown', variantAllele='G', variantType='SNP')
+class MutationField(Enum):
+    ALLELE_SPECIFIC_COPY_NUMBER = "alleleSpecificCopyNumber"
+    AMINO_ACID_CHANGE = "aminoAcidChange"
+    CENTER = "center"
+    CHROMOSOME = "chr"
+    DRIVER_FILTER = "driverFilter"
+    DRIVER_FILTER_ANNOTATION = "driverFilterAnnotation"
+    DRIVER_TIERS_FILTER = "driverTiersFilter"
+    DRIVER_TIERS_FILTER_ANNOTATION = "driverTiersFilterAnnotation"
+    END_POSITION = "endPosition"
+    KEYWORD = "keyword"
+    MUTATION_STATUS = "mutationStatus"
+    MUTATION_TYPE = "mutationType"
+    NAMESPACE_COLUMNS = "namespaceColumns"
+    NCBI_BUILD = "ncbiBuild"
+    NORMAL_ALT_COUNT = "normalAltCount"
+    NORMAL_REF_COUNT = "normalRefCount"
+    PROTEIN_CHANGE = "proteinChange"
+    PROTEIN_POS_END = "proteinPosEnd"
+    PROTEIN_POS_START = "proteinPosStart"
+    REFERENCE_ALLELE = "referenceAllele"
+    REFSEQ_MRNA_ID = "refseqMrnaId"
+    START_POSITION = "startPosition"
+    TUMOR_ALT_COUNT = "tumorAltCount"
+    TUMOR_REF_COUNT = "tumorRefCount"
+    UNIQUE_PATIENT_KEY = "uniquePatientKey"
+    UNIQUE_SAMPLE_KEY = "uniqueSampleKey"
+    VALIDATION_STATUS = "validationStatus"
+    VARIANT_ALLELE = "variantAllele"
+    VARIANT_TYPE = "variantType"
+    _ID = ["molecularProfileId", "sampleId", "patientId", "entrezGeneId", "studyId"]
+    _LABEL = "Mutation"

config/schema_config.yaml

@@ -97,6 +97,41 @@ clinical attribute:
         patientAttribute: str
         priority: str
 
+mutation:
+    is_a: molecular entity
+    represented_as: node
+    input_label: Mutation
+    properties:
+        allele_specific_copy_number: str
+        amino_acid_change: str
+        c_position: str
+        chromosome: str
+        end_position: str
+        start_position: str
+        keyword: str
+        mutation_status: str
+        mutation_type: str
+        ncbi_build: str
+        normal_alt_count: str
+        normal_ref_count: str
+        protein_change: str
+        protein_pos_start: str
+        protein_pos_end: str
+        reference_allele: str
+        variant_allele: str
+        refseq_mrna_id: str
+        tumor_alt_count: str
+        tumor_ref_count: str
+        unique_sample_key: str
+        unique_patient_key: str
+        validation_status: str
+        variant_type: str
+
+
+
+# ---------------------------------------------------
+# associations
+
 is child of:
     is_a: [association, disease disease association]
     represented_as: edge
@@ -127,12 +162,22 @@ has clinical attribute:
     represented_as: edge
     input_label: hasClinicalAttribute
 
-has gene:
+part of:
     is_a: association
     represented_as: edge
-    input_label: hasGene
+    input_label: partOf
 
-part of:
+from sample:
+    is_a: association
+    represented_as: edge
+    input_label: fromSample
+
+from gene:
+    is_a: association
+    represented_as: edge
+    input_label: fromGene
+
+from molecular profile:
     is_a: association
     represented_as: edge
-    input_label: partOf
+    input_label: fromMolecularProfile

create_knowledge_graph.py

@@ -10,29 +10,31 @@
 bc = BioCypher()
 
 node_types = [
-    PatientSampleStudyEntityField,
+    MutationField,
     MolecularProfileField,
-    SampleListField,
-    ClinicalAttributesField,
+    # SampleListField,
+    # ClinicalAttributesField,
     GeneField,
-    GenePanelField,
-    CancerTypeField,
+    # GenePanelField,
+    # CancerTypeField,
     StudyField,
     PatientField,
-    SampleField
+    SampleField,
 ]
 edge_types = [
-    PatientToPatientSampleStudyEntityField,
-    SampleToPatientSampleStudyEntityField,
-    StudyToPatientSampleStudyEntityField,
-    SampleListToStudyField,
-    StudyToClinicalDataField,
-    MolecularProfiletoStudyField,
-    GenePanelGeneAssociationField,
-    DiseaseDiseaseAssociationField,
-    StudyDiseaseAssociationField,
-    studyPatientAssociationField,
-    samplePatientAssociationField
+    # SampleListToStudyField,
+    # StudyToClinicalDataField,
+    # MolecularProfiletoStudyField,
+    # DiseaseDiseaseAssociationField,
+    # StudyDiseaseAssociationField,
+    # studyPatientAssociationField,
+    # samplePatientAssociationField,
+    mutationToMolecularProfileField,
+    mutationToSampleField,
+    mutationToGeneField,
+    mutationToStudyField,
+    mutationToPatientField
+
 ]
 
 # Create a protein adapter instance

get_api_data.py

@@ -0,0 +1,4 @@
+from bravado.client import SwaggerClient
+
+SwaggerClient.from_url('https://www.cbioportal.org/api/v2/api-docs',
+                                    config={"validate_requests":False,"validate_responses":False,"validate_swagger_spec": False})