release commit

config/biocypher_config.yaml

@@ -0,0 +1,18 @@
+# add your settings here (overriding the defaults)
+
+biocypher:
+  dbms: neo4j
+  offline: true
+  #debug: true
+  output_directory: /neo4j_import #comment if you want to debug, so that bc creates a new folder for each run in /biocypher-out
+  schema_config_path: config/automated_schema.yaml
+
+  head_ontology:
+    url: config/head_ontology/biolink-model.owl.ttl
+    root_node: entity
+
+neo4j:
+  delimiter: '\t'
+  array_delimiter: '|'
+  skip_duplicate_nodes: true
+  skip_bad_relationships: true

config/manual_schema_config.yaml

@@ -0,0 +1,283 @@
+Title: BioCypher graph schema configuration file
+
+# This configuration file establishes the hierarchy and connectivity in a newly
+# set-up BioCypher property graph database. Naming should adhere to Biolink
+# nomenclature (available at https://biolink.github.io/biolink-model/ or via 
+# the python module 'biolink-model-toolkit').
+
+# The BioCypher YAML file specifies only the leaves of the hierarchy tree of
+# the desired graph; the hierarchical structure of entities will be derived 
+# from the Biolink model + BRO model. Thus, only the immediate constituents
+# of the graph need to be specified in the schema config.
+
+
+# ---
+# "Named Things"
+# ---
+# The implementation of named things is fairly straightforward, since they are
+# usually represented in node form, which is also the Biolink recommendation.
+# The same is not true for associations.
+#
+# A little more complex is the representation of aggregates of named things.
+
+clinicalStatus:
+    is_a: ClinicalEntity 
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: clinicalStatus
+    properties:
+        coding_system: str
+        label: str
+        coding_code: str
+
+Condition:
+    is_a: ClinicalEntity 
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: Condition
+    properties:
+      input_format: HL7 FHIR
+      data_specification: Medical Informatics Initiative Germany Core Data Set, Basic Modules
+
+diagnosis:
+    is_a: ClinicalEntity
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: diagnosis
+    properties:
+        type.coding_code: str
+        sequence: str
+        label: str
+        type.coding_system: str
+
+DiagnosticReport:
+    is_a: ClinicalEntity 
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: DiagnosticReport
+    properties:
+        resourceType: str
+        label: str
+        status: str
+        id: str
+
+Encounter:
+    is_a: ClinicalEntity 
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: Encounter
+    properties:
+        resourceType: str
+        label: str
+        status: str
+        id: str
+
+identifier:
+    is_a: Attribute
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: identifier
+    properties:
+        label: str
+        value: str
+        system: str
+
+interpretation: #
+    is_a: named thing
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: interpretation
+    properties:
+        extension.valueCoding_system: str
+        extension_url: str
+        extension.valueCoding_display: str
+        coding_code: str
+        coding_system: str
+        label: str
+        extension.valueCoding_code: str
+
+maritalStatus:
+    is_a: OrganismAttribute 
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: maritalStatus
+    properties:
+        label: str
+        coding_system: str
+        coding_code: str
+
+Observation:
+    is_a: ClinicalEntity 
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: Observation
+    properties:
+        resourceType: str
+        label: str
+        effectiveDateTime: str
+        status: str
+        id: str
+
+Organization:
+    is_a: AdministrativeEntity
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: Organization
+    properties:
+        label: str
+        id: str
+        name: str
+        resourceType: str
+
+Patient:
+    is_a: Human 
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: Patient
+    properties:
+        resourceType: str
+        label: str
+        gender: str
+        id: str
+        birthDate: str
+
+Procedure:
+   # is_a: Procedure 
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: Procedure
+    properties:
+        label: str
+        performedDateTime: str
+        resourceType: str
+        status: str
+        id: str
+
+referenceRange: #
+    is_a: named thing
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: referenceRange
+    properties:
+        high_system: str
+        high_value: str
+        high_code: str
+        label: str
+        high_unit: str
+
+search: #
+    is_a: named thing
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: search
+    properties:
+        label: str
+        mode: str
+
+type:
+    is_a: Attribute
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: type
+    properties:
+        coding_system: str
+        label: str
+        coding_code: str
+        coding_display: str
+
+verificationStatus:
+    is_a: Attribute
+    represented_as: node
+    preferred_id: fhir_id
+    label_in_input: verificationStatus
+    properties:
+        coding_system: str
+        label: str
+        coding_code: str
+        coding_display: str
+
+
+# ---
+# Associations
+# ---
+# Associations are not supposed to be represented in node form as per the 
+# specifications of Biolink. However, in an analytic context, it often makes 
+# sense to represent interactions as nodes in Neo4j, because it enables, for
+# instance, the annotation of a relationship with a publication as source of
+# evidence (also known as reification in the knowledge graph world). 
+
+# The Biolink specifications for these types of relationships do 
+# not go into depth; for example, the hierarchy for molecular interactions
+# (ie, "associations") ends at "PairwiseMolecularInteraction", there are no 
+# explicit terms for protein-protein-interaction, phosphorylation, miRNA-
+# targeting, etc. Biolink proposes to use interaction identifiers from 
+# ontologies, such as https://www.ebi.ac.uk/ols/ontologies/mi/.
+
+        # association to connect anything to an identifier node
+        # if functional, includes: 
+        # IDENTIFIED_BY_Condition_Identifier, 
+        # IDENTIFIED_BY_DiagnosticReport_Identifier, 
+        # IDENTIFIED_BY_Encounter_Identifier, 
+        # IDENTIFIED_BY_Observation_Identifier, 
+        # IDENTIFIED_BY_Organization_Identifier
+        # IDENTIFIED_BY_Patient_Identifier,
+        # IDENTIFIED_BY_Procedure_Identifier
+
+condition to identifier association: 
+  is_a: association
+  represented_as: edge
+  label_in_input: IDENTIFIED_BY_Condition_Identifier
+
+diagnostic report to identifier association: 
+  is_a: association
+  represented_as: edge
+  label_in_input: IDENTIFIED_BY_DiagnosticReport_Identifier
+
+observation to identifier association: 
+  is_a: association
+  represented_as: edge
+  label_in_input: IDENTIFIED_BY_Observation_Identifier
+
+observation derived from observation association: 
+  is_a: association
+  represented_as: edge
+  label_in_input: DERIVED_FROM_Observation_Observation
+
+observation has member observation association: 
+  is_a: association
+  represented_as: edge
+  label_in_input: HAS_MEMBER_Observation_Observation
+
+procedure to identifier association: 
+  is_a: association
+  represented_as: edge
+  label_in_input: IDENTIFIED_BY_Procedure_Identifier
+
+procedure to diagnostic report association: 
+  is_a: association
+  represented_as: edge
+  label_in_input: IDENTIFIED_BY_Procedure_Identifier
+
+procedure reasoned by observation association: 
+  is_a: association
+  represented_as: edge
+  label_in_input: HAS_REASON_REFERENCE_Procedure_Observation
+
+procedure performer is practitioner association:
+  is_a: association
+  represented_as: edge
+  label_in_input: HAS_ACTOR_ProcedurePerformer_Practitioner
+
+  #represented_as: edge
+  #label_in_input: DERIVED_FROM_Observation_Observation:
+  #represented_as: edge
+  #label_in_input: DERIVED_FROM_Observation_Observation
+      #protein interaction:
+      #    is_a: Pairwise molecular interaction
+      #    represented_as: edge
+      #    label_in_input: protein_protein_interaction
+
+      #protein to disease association:
+      #   is_a: Association
+      #    represented_as: edge
+      #    label_in_input: protein_disease_association