def meta_node_df(meta_node):
query = META_NODE_QUERY_TEMPLATE.format(meta_node=meta_node)
driver = neo4j_connect()
session = driver.session()
data = session.run(query).data()
df = pd.json_normalize(data)
return df
def query_graph(query):
# collect to epigraph
driver = neo4j_connect()
session = driver.session()
# query
query_data = session.run(query).data()
df = pd.json_normalize(query_data)
return df
def test_connect():
query = """
MATCH (n)
RETURN n LIMIT 2;
"""
driver = neo4j_connect()
session = driver.session()
data = session.run(query).data()
logger.info(data)
assert len(data) == 2
def get_disease_data():
driver = neo4j_connect()
session = driver.session()
query = """
match (d:Disease) unwind(d.efo) as mondo_efo_id return d.id as disease_id, mondo_efo_id;
"""
query_data = session.run(query).data()
df = pd.json_normalize(query_data)
logger.info(df)
return df
def test_meta_node_index():
query = "CALL db.indexes()"
driver = neo4j_connect()
with driver.session() as session:
data = session.run(query).data()
actual_node_indexes = {(_["labelsOrTypes"][0], _["properties"][0]) for _ in data}
expected_node_indexes = {
(key, value["index"]) for key, value in meta_node_dict.items()
}
# We would expect expected_indexes to be a subset of current_indexes
assert len(expected_node_indexes.difference(actual_node_indexes)) == 0
def source_target_dict(meta_rel):
query = SOURCE_TARGET_QUERY_TEMPLATE.format(meta_rel=meta_rel)
driver = neo4j_connect()
session = driver.session()
data = session.run(query).data()
source = [_["source"] for _ in data]
target = [_["target"] for _ in data]
res = {
"source": [_ for sub_list in source for _ in sub_list],
"target": [_ for sub_list in target for _ in sub_list],
}
return res
def test_meta_rel_exist():
query = """
CALL db.relationshipTypes()
YIELD relationshipType
RETURN relationshipType AS label
"""
driver = neo4j_connect()
with driver.session() as session:
data = session.run(query).data()
db_meta_rel_names = set([_["label"] for _ in data])
logger.info(f"meta_rel_names: {meta_rel_names}")
logger.info(f"db_meta_rel_names: {db_meta_rel_names}")
assert set(meta_rel_names) == db_meta_rel_names
def test_meta_node_exist():
query = """
CALL db.labels() YIELD label RETURN label
"""
driver = neo4j_connect()
with driver.session() as session:
data = session.run(query).data()
db_meta_node_names = set([_["label"] for _ in data])
if {"Meta"}.issubset(db_meta_node_names):
db_meta_node_names.remove({"Meta"})
logger.info(f"meta_node_names: {meta_node_names}")
logger.info(f"db_meta_node_names: {db_meta_node_names}")
assert set(meta_node_names) == db_meta_node_names
def get_variants_from_graph():
# collect to epigraph
driver = neo4j_connect()
session = driver.session()
# query
query = """
match (v:Variant)
return distinct(v._id) as id limit 100
"""
logger.info(query)
query_data = session.run(query).data()
df = pd.json_normalize(query_data)
df.to_csv(variant_data, index=False)
copy_source_data(data_name=data_name, filename=variant_data)
return df
def check():
driver = neo4j_connect()
session = driver.session()
# read data
data = os.path.join(dataDir, FILE)
df = pd.read_csv(data, sep="\t")
print(df.head())
ens_list = list(set(list(df["ensembl_gene_id"])))
print(len(ens_list))
com = """
match (g:Gene) where g.ensembl_id in {ens_list} return g.ensembl_id;
""".format(ens_list=ens_list)
# print(com)
result = session.run(com)
res_df = pd.DataFrame([dict(record) for record in result])
print(res_df)
# find missing
print(set(list(df["ensembl_gene_id"])) - set(list(res_df["g.ensembl_id"])))
