Merge pull request gyorilab#344 from nanglo123/add_dkg_resource

Add resources to dkg

docker/Dockerfile

@@ -31,7 +31,7 @@ RUN wget -O /sw/nodes.tsv.gz https://askem-mira.s3.amazonaws.com/dkg/$domain/bui
 
 # Python packages
 RUN python -m pip install --upgrade pip && \
- python -m pip install git+https://github.com/gyorilab/mira.git@main#egg=mira[web,uvicorn,dkg-client] && \
+ python -m pip install git+https://github.com/gyorilab/mira.git@main#egg=mira[web,uvicorn,dkg-client,dkg-construct] && \
  python -m pip uninstall -y flask_bootstrap && \
  python -m pip uninstall -y bootstrap_flask && \
  python -m pip install bootstrap_flask && \

mira/dkg/api.py

@@ -2,7 +2,7 @@
 
 import itertools as itt
 import os
-from typing import Any, List, Mapping, Optional, Union, Dict
+from typing import Any, List, Mapping, Optional, Union
 
 import pydantic
 from fastapi import APIRouter, Body, HTTPException, Path, Query, Request
@@ -13,6 +13,7 @@
 
 from mira.dkg.client import AskemEntity, Entity, Relation
 from mira.dkg.utils import DKG_REFINER_RELS
+from mira.dkg.construct import add_resource_to_dkg
 
 __all__ = [
  "api_blueprint",
@@ -360,6 +361,37 @@ def add_relations(
  request.app.state.client.add_relation(relation)
 
 
+ @api_blueprint.post(
+ "/add_resources",
+ response_model=None,
+ tags=["relations"],
+ )
+ def add_resources(
+ request: Request,
+ resource_prefix_list: List[str] = Body(
+ ...,
+ description="A of resources to add to the DKG",
+ title="Resource Prefixes",
+ example=["probonto", "wikidata", "eiffel", "geonames", "ncit",
+ "nbcbitaxon"],
+ )
+ ):
+ """From a list of resource prefixes, add a list of nodes and edges
+ extract from each resource to the DKG"""
+ for resource_prefix in resource_prefix_list:
+ # nodes and edges will be a list of dicts
+ nodes, edges = add_resource_to_dkg(resource_prefix.lower())
+ # node_info and edge_info are dictionaries that will be
+ # unpacked when creating instances of entities and relations
+ entities = [Entity(**node_info) for node_info in nodes]
+ relations = [Relation(**edge_info) for edge_info in edges]
+
+ for entity in entities:
+ request.app.state.client.add_node(entity)
+ for relation in relations:
+ request.app.state.client.add_relation(relation)
+
+
 class IsOntChildResult(BaseModel):
  """Result of a query to /is_ontological_child"""
 

mira/dkg/client.py

@@ -170,6 +170,7 @@ def from_data(cls, data):
  -------
  A MIRA entity
  """
+
  if isinstance(data, neo4j.graph.Node):
  data = dict(data.items())
  properties = defaultdict(list)
@@ -185,6 +186,8 @@ def from_data(cls, data):
  ):
  synonyms.append(Synonym(value=value, type=type))
  xrefs = []
+
+
  for curie, type in zip(
  data.pop("xrefs", []),
  data.pop("xref_types", []),
@@ -343,29 +346,70 @@ def add_node(self, entity):
  entity:
  The node object that will be added to the DKG
  """
- curie = entity.id
+ xrefs, xref_types = [], []
+ synonyms, synonym_types = [], []
+ property_predicates, property_values = [], []
+ for xref in entity.xrefs:
+ xrefs.append(xref.id)
+ xref_types.append(xref.type)
+ for synonym in entity.synonyms:
+ synonyms.append(synonym.value)
+ synonym_types.append(synonym.type)
+ for property_predicate, property_value_list in entity.properties.items():
+ property_predicates.append(property_predicate)
+ property_values.extend(property_value_list)
+
+ _id = entity.id
  name = entity.name
  type = entity.type
  obsolete = entity.obsolete
  description = entity.description
- synonyms = entity.synonyms
  alts = entity.alts
- xrefs = entity.xrefs
  labels = entity.labels
 
  create_source_node_query = (
- f"MERGE (n {{curie: '{curie}', "
- f"name: '{name}', "
- f"type: '{type}', "
- f"obsolete: {obsolete}, "
- f"description: '{description}', "
- f"synonyms: {synonyms}, "
- f"alts: {alts}, "
- f"xrefs: {xrefs}, "
- f"labels: {labels} }} )"
+ "MERGE (n {id: $id, "
+ "type: $type, "
+ "obsolete: $obsolete"
  )
 
- self.create_tx(create_source_node_query)
+ if name:
+ create_source_node_query += ", name: $name"
+ if description:
+ create_source_node_query += ", description: $description"
+ if alts:
+ create_source_node_query += ", alts: $alts"
+ if labels:
+ create_source_node_query += ", labels: $labels"
+ if xrefs:
+ create_source_node_query += ", xrefs: $xrefs"
+ create_source_node_query += ", xref_types: $xref_types"
+ if synonyms:
+ create_source_node_query += ", synonyms: $synonyms"
+ create_source_node_query += ", synonym_types: $synonym_types"
+ if property_predicates:
+ create_source_node_query += ", property_predicates: $property_predicates"
+ create_source_node_query += ", property_values: $property_values"
+
+ create_source_node_query += "})"
+
+ query_parameters = {
+ "id": _id,
+ "name": name,
+ "type": type,
+ "obsolete": obsolete,
+ "description": description,
+ "synonyms": synonyms,
+ "synonym_types": synonym_types,
+ "alts": alts,
+ "xrefs": xrefs,
+ "xref_types": xref_types,
+ "labels": labels,
+ "property_predicates": property_predicates,
+ "property_values": property_values
+ }
+
+ self.create_tx(create_source_node_query, **query_parameters)
 
  def add_relation(self, relation):
  """Add a relation to the DKG
@@ -384,8 +428,8 @@ def add_relation(self, relation):
  graph = relation.graph
 
  create_relation_query = (
- f"MATCH (source_node {{curie: '{source_curie}'}}), "
- f"(target_node {{curie: '{target_curie}'}}) "
+ f"MATCH (source_node {{id: '{source_curie}'}}), "
+ f"(target_node {{id: '{target_curie}'}}) "
  f"MERGE (source_node)-[rel:{type}]->(target_node)"
  f"SET rel.pred = '{pred}'"
  f"SET rel.source = '{source}'"
@@ -395,7 +439,6 @@ def add_relation(self, relation):
 
  self.create_tx(create_relation_query)
 
-
  def create_single_property_node_index(
  self,
  index_name: str,