def get_subgraph(seedNodes): # noqa: E501
"""Get a GNBR subgraph
Query a list of nodes to get the GNBR subgraph induced by those nodes # noqa: E501
:param seedNodes: gnbrIDs for subgraph
:type seedNodes: List[str]
:rtype: GnbrSubgraph
"""
q = """
MATCH (n) WHERE n.id IN {seedNodes}
MATCH p=(n)--(m)
WHERE m.id IN {seedNodes}
RETURN relationships(p)
LIMIT 25
"""
results = neo4j.run(query=q, param={"seedNodes": seedNodes})
subgraph = GnbrSubgraph()
for d in results:
gnbr_edge = GNBR_edge()
## Add some subgraph!!!
# TODO!!
return subgraph
def get_node_neighbors(entity1): # noqa: E501
"""Get all neighbors of a particular node
Query node to get all nodes connected by at least one edge to the input node within GNBR # noqa: E501
:param entity1: GNBR-ID for first entity
:type entity1: str
:rtype: GnbrSubgraph
"""
q = """
MATCH p=(n)--(m)
WHERE m.id = {entity1}
RETURN relationships(p)
LIMIT 25
"""
results = neo4j.run(query=q, param={"entity1": entity1})
subgraph = GnbrSubgraph()
for d in results:
gnbr_edge = GNBR_edge()
## Add some subgraph!!!
# TODO!!
return subgraph
def get_concept_details(conceptId):
"""
get_concept_details
Retrieves details for a specified concepts in the system, as specified by a (url-encoded) CURIE identifier of a concept known the given knowledge source.
:param conceptId: (url-encoded) CURIE identifier of concept of interest
:type conceptId: str
:rtype: List[ConceptDetail]
"""
q = """
MATCH (protein:Protein {stringId: {conceptId}})
RETURN
protein.stringId as stringId,
protein.alias as alias,
protein.source as source
LIMIT 1
"""
results = neo4j.run(query=q, param={"conceptId" : conceptId})
details = []
for d in results:
concept_detail = ConceptDetail()
concept_detail.id = d["stringId"]
concept_detail.name = d["alias"]
concept_detail.semantic_group = "CHEM"
concept_detail.details = [ConceptsconceptIdDetails(tag="source", value=d["source"])]
details.append(concept_detail)
return details
def linked_types():
"""
linked_types
Get a list of types and # of instances in the knowledge source, and a link to the API call for the list of equivalent terminology
:rtype: List[DataType]
"""
results = neo4j.run("""
MATCH (p:Protein)
RETURN COUNT(p) as frequency
""")
return [DataType(id="CHEM", frequency=d["frequency"]) for d in results]
def get_concepts(keywords, semgroups=None, pageNumber=None, pageSize=None):
"""
get_concepts
Retrieves a (paged) list of concepts in the system
:param keywords: a (urlencoded) space delimited set of keywords or substrings against which to match concept names and synonyms
:type keywords: str
:param semgroups: a (url-encoded) space-delimited set of semantic groups (specified as codes CHEM, GENE, ANAT, etc.) to which to constrain concepts matched by the main keyword search (see [SemGroups](https://metamap.nlm.nih.gov/Docs/SemGroups_2013.txt) for the full list of codes)
:type semgroups: str
:param pageNumber: (1-based) number of the page to be returned in a paged set of query results
:type pageNumber: int
:param pageSize: number of concepts per page to be returned in a paged set of query results
:type pageSize: int
:rtype: List[Concept]
"""
q = """
MATCH (protein:Protein)
WITH
SIZE(FILTER(x IN {filter} WHERE LOWER(protein.alias) CONTAINS LOWER(x))) AS num_matches,
SIZE((protein)-[:ACTION]-()) as degree,
protein as protein
WHERE num_matches > 0 AND degree > 0
RETURN
protein.stringId as stringId,
protein.alias as alias,
protein.source as source
ORDER BY num_matches DESC, degree DESC
SKIP ({pageNumber} - 1) * {pageSize} LIMIT {pageSize}
"""
results = neo4j.run(
query=q,
param={
"pageNumber" : pageNumber if pageNumber != None and pageNumber > 0 else 1,
"pageSize" : pageSize if pageSize != None and pageSize > 0 else 10,
"filter" : keywords.split()
}
)
return [Concept(id=d["stringId"], name=d["alias"], semantic_group="CHEM") for d in results]
def get_identifier(searchString, limit=None): # noqa: E501
"""Find GNBR identifier
Searches entities within GNBR for a matching ID, based on input string # noqa: E501
:param searchString: pass a search string to find matching identifiers
:type searchString: str
:param limit: maximum number of records to return
:type limit: int
:rtype: IdMapping
"""
###### NOTE: THIS INFORMATION SHOULD PROBABLY BE FOUND USING THE PUBTATOR DICTIONARIES (SIMPLE LOOKUP) ########
q = """
MATCH (m)
WHERE m.formatted contains {searchString}
RETURN (m)
LIMIT 25
"""
results = neo4j.run(query=q, param={"searchString": searchString})
subgraph = GnbrSubgraph()
for d in results:
id_mapping = IdMapping()
## Add some subgraph!!!
# TODO!!
"""
example node output looks like:
{
"raw": "ace|acetylcholinesterase|acetylcholinesterase enzyme|ach-e|ache|ache-r|anti-acetylcholinesterase|as-ache|mache",
"formatted": "ace|acetylcholinesterase|acetylcholinesterase_enzyme|ach-e|ache|ache-r|anti-acetylcholinesterase|as-ache|mache",
"id": "11423"
}
"""
return subgraph
def get_statements(c,
pageNumber=None,
pageSize=None,
keywords=None,
semgroups=None):
"""
get_statements
Given a list of [CURIE-encoded](https://www.w3.org/TR/curie/) identifiers of exactly matching concepts, retrieves a paged list of concept-relations where either the subject or object concept matches at least one concept in the input list
:param c: set of [CURIE-encoded](https://www.w3.org/TR/curie/) identifiers of exactly matching concepts to be used in a search for associated concept-relation statements
:type c: List[str]
:param pageNumber: (1-based) number of the page to be returned in a paged set of query results
:type pageNumber: int
:param pageSize: number of concepts per page to be returned in a paged set of query results
:type pageSize: int
:param keywords: a (url-encoded, space-delimited) string of keywords or substrings against which to match the subject, predicate or object names of the set of concept-relations matched by any of the input exact matching concepts
:type keywords: str
:param semgroups: a (url-encoded, space-delimited) string of semantic groups (specified as codes CHEM, GENE, ANAT, etc.) to which to constrain the subject or object concepts associated with the query seed concept (see [SemGroups](https://metamap.nlm.nih.gov/Docs/SemGroups_2013.txt) for the full list of codes)
:type semgroups: str
:rtype: List[Statement]
"""
query = """
MATCH (a:Protein)-[r:ACTION]-(b:Protein)
WHERE
(NOT a.alias IS NULL) AND
(NOT b.alias IS NULL) AND
ANY (x in {conceptIds} WHERE
LOWER(a.stringId) = LOWER(x)
OR
LOWER(a.stringId) = LOWER(x)
)
RETURN
a.stringId as id_a,
a.alias as alias_a,
b.stringId as id_b,
b.alias as alias_b,
r.mode as relation,
ID(r) as relation_id
SKIP ({pageNumber} - 1) * {pageSize} LIMIT {pageSize}
"""
results = neo4j.run(
query, {
"pageNumber":
pageNumber if pageNumber != None and pageNumber > 0 else 1,
"pageSize": pageSize if pageSize != None and pageSize > 0 else 10,
"conceptIds": c if c != None else []
})
statements = []
for row in results:
statement = Statement()
statement_object = StatementObject()
statement_subject = StatementSubject()
statement_predicate = StatementPredicate()
statement_object.name = row["alias_a"]
statement_object.id = row["id_a"]
statement_subject.name = row["alias_b"]
statement_subject.id = row["id_b"]
statement_predicate.name = row["relation"]
statement_predicate.id = str(row["relation_id"])
statement.subject = statement_subject
statement.object = statement_object
statement.predicate = statement_predicate
statement.id = str(row["relation_id"])
statements.append(statement)
return statements
def get_edge(entity1, entity2): # noqa: E501
"""Query for an edge
Query for edges connecting two entities within GNBR # noqa: E501
:param entity1: GNBR-ID for first entity
:type entity1: str
:param entity2: GNBR-ID for second entity
:type entity2: str
:rtype: GnbrEdge
"""
q = """
MATCH p=(m)--(n)
WHERE m.id={entity1} AND n.id={entity2}
RETURN relationships(p)
LIMIT 1
"""
results = neo4j.run(query=q,
param={
"entity1": entity1,
"entity2": entity2
})
gnbr_edge = GNBR_edge()
# do something with results[0]
######
"""
Typical d in result would look like:
{
"first_entity_name_loc_char": "1428,1431",
"a-.ind": 0,
"first_entity_type": "Chemical",
"o.ind": 0,
"second_entity_name_raw": "Myc",
"sentence_number": "9",
"e+.ind": 0,
"first_entity_name": "TSA",
"dependency_path": "presence|nmod|START_ENTITY END_ENTITY|nmod|presence",
"a+.ind": 0,
"b": 164,
"first_entity_name_raw": "TSA",
"e": 639,
"e+": 56,
"k.ind": 0,
"z.ind": 0,
"b.ind": 0,
"e-": 66,
"a+": 47,
"k": 536,
"a-": 38,
"second_entity_type": "Gene",
"tax_id": "9606",
"n": 100,
"o": 154,
"second_entity_name_loc_char": "1405,1408",
"e-.ind": 0,
"e.ind": 0,
"sentence_tokenized": "TSA treatment induced a similar epidermal phenotype to activation of Myc , and activation of Myc in the presence of TSA resulted in massive stimulation of terminal differentiation .",
"z": 18,
"n.ind": 0,
"PMID": "17712411",
"second_entity_name": "Myc"
}
"""
return gnbr_edge