def test_clique_merge():
"""
Test for clique merge.
"""
input_args = {
"filename": [
os.path.join(RESOURCE_DIR, "cm_nodes.csv"),
os.path.join(RESOURCE_DIR, "cm_edges.csv"),
],
"format": "csv",
}
t = Transformer()
t.transform(input_args)
updated_graph, clique_graph = clique_merge(
target_graph=t.store.graph, prefix_prioritization_map=prefix_prioritization_map
)
leaders = NxGraph.get_node_attributes(updated_graph, "clique_leader")
leader_list = list(leaders.keys())
leader_list.sort()
assert len(leader_list) == 2
n1 = updated_graph.nodes()[leader_list[0]]
assert n1["election_strategy"] == "PREFIX_PRIORITIZATION"
assert "NCBIGene:100302240" in n1["same_as"]
assert "ENSEMBL:ENSG00000284458" in n1["same_as"]
n2 = updated_graph.nodes()[leader_list[1]]
assert n2["election_strategy"] == "PREFIX_PRIORITIZATION"
assert "NCBIGene:8202" in n2["same_as"]
assert "OMIM:601937" in n2["same_as"]
assert "ENSEMBL:ENSG00000124151" not in n2["same_as"]
def test_clique_merge():
"""
Test for clique merge.
"""
input_args = {
'filename': [
os.path.join(RESOURCE_DIR, 'cm_nodes.csv'),
os.path.join(RESOURCE_DIR, 'cm_edges.csv'),
],
'format':
'csv',
}
t = Transformer()
t.transform(input_args)
updated_graph, clique_graph = clique_merge(
target_graph=t.store.graph,
prefix_prioritization_map=prefix_prioritization_map)
leaders = NxGraph.get_node_attributes(updated_graph, 'clique_leader')
leader_list = list(leaders.keys())
leader_list.sort()
assert len(leader_list) == 2
n1 = updated_graph.nodes()[leader_list[0]]
assert n1['election_strategy'] == 'PREFIX_PRIORITIZATION'
assert 'NCBIGene:100302240' in n1['same_as']
assert 'ENSEMBL:ENSG00000284458' in n1['same_as']
n2 = updated_graph.nodes()[leader_list[1]]
assert n2['election_strategy'] == 'PREFIX_PRIORITIZATION'
assert 'NCBIGene:8202' in n2['same_as']
assert 'OMIM:601937' in n2['same_as']
assert 'ENSEMBL:ENSG00000124151' not in n2['same_as']
def test_clique_merge8():
"""
Test for clique merge where same_as appear as both node and edge properties.
"""
ppm = {"biolink:Gene": ["HGNC", "NCBIGene", "ENSEMBL", "OMIM"]}
g1 = NxGraph()
g1.add_node("HGNC:1", **{"category": ["biolink:Gene"]})
g1.add_node("OMIM:2", **{"category": ["biolink:Gene"], "same_as": ["HGNC:1"]})
g1.add_node("NCBIGene:3", **{"category": ["biolink:NamedThing"]})
g1.add_node("ENSEMBL:4", **{"category": ["biolink:Gene"], "same_as": ["HGNC:1"]})
g1.add_node(
"ENSEMBL:6", **{"category": ["biolink:Gene"], "same_as": ["NCBIGene:8"]}
)
g1.add_node("HGNC:7", **{"category": ["biolink:Gene"]})
g1.add_node("NCBIGene:8", **{"category": ["biolink:Gene"]})
g1.add_edge(
"NCBIGene:3",
"HGNC:1",
edge_key=generate_edge_key("NCBIGene:3", "biolink:same_as", "HGNC:1"),
**{"predicate": "biolink:same_as", "relation": "owl:equivalentClass"}
)
g1.add_edge(
"ENSEMBL:6",
"NCBIGene:8",
edge_key=generate_edge_key("ENSEMBL:6", "biolink:same_as", "NCBIGene:8"),
**{"predicate": "biolink:same_as", "relation": "owl:equivalentClass"}
)
g1.add_edge(
"HGNC:7",
"NCBIGene:8",
edge_key=generate_edge_key("HGNC:7", "biolink:same_as", "NCBIGene:8"),
**{"predicate": "biolink:same_as", "relation": "owl:equivalentClass"}
)
updated_graph, clique_graph = clique_merge(
target_graph=g1, prefix_prioritization_map=ppm
)
assert updated_graph.number_of_nodes() == 2
assert updated_graph.number_of_edges() == 0
assert updated_graph.has_node("HGNC:1")
assert updated_graph.has_node("HGNC:7")
n1 = updated_graph.nodes()["HGNC:1"]
assert "OMIM:2" in n1["same_as"]
assert "NCBIGene:3" in n1["same_as"]
assert "ENSEMBL:4" in n1["same_as"]
n2 = updated_graph.nodes()["HGNC:7"]
assert "ENSEMBL:6" in n2["same_as"]
assert "NCBIGene:8" in n2["same_as"]
assert not updated_graph.has_node("OMIM:2")
assert not updated_graph.has_node("NCBIGene:3")
assert not updated_graph.has_node("ENSEMBL:4")
assert not updated_graph.has_node("ENSEMBL:6")
assert not updated_graph.has_node("NCBIGene:8")
def test_clique_generation():
"""
Test for generation of cliques.
"""
input_args = {
"filename": [
os.path.join(RESOURCE_DIR, "cm_nodes.csv"),
os.path.join(RESOURCE_DIR, "cm_edges.csv"),
],
"format": "csv",
}
t = Transformer()
t.transform(input_args)
updated_graph, clique_graph = clique_merge(
target_graph=t.store.graph, prefix_prioritization_map=prefix_prioritization_map
)
cliques = list(nx.strongly_connected_components(clique_graph))
assert len(cliques) == 2
def test_clique_merge_edge_consolidation():
"""
Test for clique merge, with edge consolidation.
"""
input_args = {
'filename': [
os.path.join(RESOURCE_DIR, 'cm_test2_nodes.tsv'),
os.path.join(RESOURCE_DIR, 'cm_test2_edges.tsv'),
],
'format':
'tsv',
}
t = Transformer()
t.transform(input_args)
updated_graph, clique_graph = clique_merge(
target_graph=t.store.graph,
prefix_prioritization_map=prefix_prioritization_map)
leaders = NxGraph.get_node_attributes(updated_graph, 'clique_leader')
leader_list = list(leaders.keys())
leader_list.sort()
assert len(leader_list) == 2
n1 = updated_graph.nodes()[leader_list[0]]
assert n1['election_strategy'] == 'LEADER_ANNOTATION'
assert 'NCBIGene:100302240' in n1['same_as']
assert 'ENSEMBL:ENSG00000284458' in n1['same_as']
n2 = updated_graph.nodes()[leader_list[1]]
assert n2['election_strategy'] == 'LEADER_ANNOTATION'
assert 'NCBIGene:8202' in n2['same_as']
assert 'OMIM:601937' in n2['same_as']
assert 'ENSEMBL:ENSG00000124151' not in n2['same_as']
e1_incoming = updated_graph.in_edges('HGNC:7670', data=True)
assert len(e1_incoming) == 3
e1_outgoing = updated_graph.out_edges('HGNC:7670', data=True)
assert len(e1_outgoing) == 6
def test_clique_merge_edge_consolidation():
"""
Test for clique merge, with edge consolidation.
"""
input_args = {
"filename": [
os.path.join(RESOURCE_DIR, "cm_test2_nodes.tsv"),
os.path.join(RESOURCE_DIR, "cm_test2_edges.tsv"),
],
"format": "tsv",
}
t = Transformer()
t.transform(input_args)
updated_graph, clique_graph = clique_merge(
target_graph=t.store.graph, prefix_prioritization_map=prefix_prioritization_map
)
leaders = NxGraph.get_node_attributes(updated_graph, "clique_leader")
leader_list = list(leaders.keys())
leader_list.sort()
assert len(leader_list) == 2
n1 = updated_graph.nodes()[leader_list[0]]
assert n1["election_strategy"] == "LEADER_ANNOTATION"
assert "NCBIGene:100302240" in n1["same_as"]
assert "ENSEMBL:ENSG00000284458" in n1["same_as"]
n2 = updated_graph.nodes()[leader_list[1]]
assert n2["election_strategy"] == "LEADER_ANNOTATION"
assert "NCBIGene:8202" in n2["same_as"]
assert "OMIM:601937" in n2["same_as"]
assert "ENSEMBL:ENSG00000124151" not in n2["same_as"]
e1_incoming = updated_graph.in_edges("HGNC:7670", data=True)
assert len(e1_incoming) == 3
e1_outgoing = updated_graph.out_edges("HGNC:7670", data=True)
assert len(e1_outgoing) == 6
def test_clique_merge1():
"""
Test to perform a clique merge where all nodes in a clique are valid.
"""
ppm = {'biolink:Gene': ['HGNC', 'NCBIGene', 'ENSEMBL', 'OMIM']}
g1 = NxGraph()
g1.add_node('HGNC:1', **{'category': ['biolink:Gene']})
g1.add_node('OMIM:2', **{'category': ['biolink:Gene']})
g1.add_node('NCBIGene:3', **{'category': ['biolink:Gene']})
g1.add_node('ENSEMBL:4', **{'category': ['biolink:Gene']})
g1.add_node('ENSEMBL:6', **{'category': ['biolink:Gene']})
g1.add_node('HGNC:7', **{'category': ['biolink:Gene']})
g1.add_node('NCBIGene:8', **{'category': ['biolink:Gene']})
g1.add_edge('ENSEMBL:4',
'HGNC:1',
edge_key=generate_edge_key('ENSEMBL:4', 'biolink:same_as',
'HGNC:1'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('NCBIGene:3',
'HGNC:1',
edge_key=generate_edge_key('NCBIGene:3', 'biolink:same_as',
'HGNC:1'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('OMIM:2',
'HGNC:1',
edge_key=generate_edge_key('OMIM:2', 'biolink:same_as',
'HGNC:1'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('ENSEMBL:6',
'NCBIGene:8',
edge_key=generate_edge_key('ENSEMBL:6', 'biolink:same_as',
'NCBIGene:8'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('HGNC:7',
'NCBIGene:8',
edge_key=generate_edge_key('HGNC:7', 'biolink:same_as',
'NCBIGene:8'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
updated_graph, clique_graph = clique_merge(target_graph=g1,
prefix_prioritization_map=ppm)
print_graph(updated_graph)
assert updated_graph.number_of_nodes() == 2
assert updated_graph.number_of_edges() == 0
assert updated_graph.has_node('HGNC:1')
assert updated_graph.has_node('HGNC:7')
n1 = updated_graph.nodes()['HGNC:1']
assert 'OMIM:2' in n1['same_as']
assert 'NCBIGene:3' in n1['same_as']
assert 'ENSEMBL:4' in n1['same_as']
n2 = updated_graph.nodes()['HGNC:7']
assert 'ENSEMBL:6' in n2['same_as']
assert 'NCBIGene:8' in n2['same_as']
assert not updated_graph.has_node('OMIM:2')
assert not updated_graph.has_node('NCBIGene:3')
assert not updated_graph.has_node('ENSEMBL:4')
assert not updated_graph.has_node('ENSEMBL:6')
assert not updated_graph.has_node('NCBIGene:8')
def test_clique_merge9():
"""
Test for clique merge where same_as appear as both node and edge properties,
but an invalid node also has a same_as property and participates in same_as edge.
"""
ppm = {'biolink:Gene': ['HGNC', 'NCBIGene', 'ENSEMBL', 'OMIM']}
g1 = NxGraph()
g1.add_node('HGNC:1', **{'category': ['biolink:Gene']})
g1.add_node('OMIM:2', **{
'category': ['biolink:Disease'],
'same_as': ['HGNC:1']
})
g1.add_node('NCBIGene:3', **{'category': ['biolink:NamedThing']})
g1.add_node('ENSEMBL:4', **{
'category': ['biolink:Gene'],
'same_as': ['HGNC:1']
})
g1.add_node('ENSEMBL:6', **{
'category': ['biolink:Gene'],
'same_as': ['NCBIGene:8']
})
g1.add_node('HGNC:7', **{'category': ['biolink:Gene']})
g1.add_node('NCBIGene:8', **{'category': ['biolink:Gene']})
g1.add_edge('X:00001',
'OMIM:2',
edge_key=generate_edge_key('X:00001', 'biolink:same_as',
'OMIM:2'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('NCBIGene:3',
'HGNC:1',
edge_key=generate_edge_key('NCBIGene:3', 'biolink:same_as',
'HGNC:1'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('ENSEMBL:6',
'NCBIGene:8',
edge_key=generate_edge_key('ENSEMBL:6', 'biolink:same_as',
'NCBIGene:8'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('HGNC:7',
'NCBIGene:8',
edge_key=generate_edge_key('HGNC:7', 'biolink:same_as',
'NCBIGene:8'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
updated_graph, clique_graph = clique_merge(target_graph=g1,
prefix_prioritization_map=ppm)
assert updated_graph.number_of_nodes() == 4
assert updated_graph.number_of_edges() == 1
assert updated_graph.has_node('HGNC:1')
assert updated_graph.has_node('HGNC:7')
n1 = updated_graph.nodes()['HGNC:1']
assert 'OMIM:2' not in n1['same_as']
assert 'NCBIGene:3' in n1['same_as']
assert 'ENSEMBL:4' in n1['same_as']
n2 = updated_graph.nodes()['HGNC:7']
assert 'ENSEMBL:6' in n2['same_as']
assert 'NCBIGene:8' in n2['same_as']
assert updated_graph.has_node('OMIM:2')
def test_clique_merge7():
"""
Test for clique merge where each clique has a node that has
a disjoint category from other nodes in a clique and the node is
not a participant in same_as edges.
"""
ppm = {'biolink:Gene': ['HGNC', 'NCBIGene', 'ENSEMBL', 'OMIM']}
g1 = NxGraph()
g1.add_node('HGNC:1', **{'category': ['biolink:Gene']})
g1.add_node('OMIM:2', **{'category': ['biolink:Disease']})
g1.add_node('NCBIGene:3', **{'category': ['biolink:NamedThing']})
g1.add_node('ENSEMBL:4', **{'category': ['biolink:Gene']})
g1.add_node('ENSEMBL:6', **{'category': ['biolink:Gene']})
g1.add_node('HGNC:7', **{'category': ['biolink:Disease']})
g1.add_node('NCBIGene:8', **{'category': ['biolink:Gene']})
g1.add_edge('ENSEMBL:4',
'HGNC:1',
edge_key=generate_edge_key('ENSEMBL:4', 'biolink:same_as',
'HGNC:1'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('NCBIGene:3',
'HGNC:1',
edge_key=generate_edge_key('NCBIGene:3', 'biolink:same_as',
'HGNC:1'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('OMIM:2',
'HGNC:1',
edge_key=generate_edge_key('OMIM:2', 'biolink:same_as',
'HGNC:1'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('ENSEMBL:6',
'NCBIGene:8',
edge_key=generate_edge_key('ENSEMBL:6', 'biolink:same_as',
'NCBIGene:8'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
g1.add_edge('HGNC:7',
'NCBIGene:8',
edge_key=generate_edge_key('HGNC:7', 'biolink:same_as',
'NCBIGene:8'),
**{
'predicate': 'biolink:same_as',
'relation': 'owl:equivalentClass'
})
updated_graph, clique_graph = clique_merge(target_graph=g1,
prefix_prioritization_map=ppm)
assert updated_graph.number_of_nodes() == 4
assert updated_graph.number_of_edges() == 2
assert updated_graph.has_node('HGNC:1')
assert updated_graph.has_node('NCBIGene:8')
n1 = updated_graph.nodes()['HGNC:1']
assert 'NCBIGene:3' in n1['same_as']
assert 'ENSEMBL:4' in n1['same_as']
assert 'OMIM:2' not in n1['same_as']
n2 = updated_graph.nodes()['NCBIGene:8']
assert 'ENSEMBL:6' in n2['same_as']
assert updated_graph.has_node('OMIM:2')
assert not updated_graph.has_node('NCBIGene:3')
assert not updated_graph.has_node('ENSEMBL:4')
assert updated_graph.has_node('HGNC:7')
def test_clique_merge7():
"""
Test for clique merge where each clique has a node that has
a disjoint category from other nodes in a clique and the node is
not a participant in same_as edges.
"""
ppm = {"biolink:Gene": ["HGNC", "NCBIGene", "ENSEMBL", "OMIM"]}
g1 = NxGraph()
g1.add_node("HGNC:1", **{"category": ["biolink:Gene"]})
g1.add_node("OMIM:2", **{"category": ["biolink:Disease"]})
g1.add_node("NCBIGene:3", **{"category": ["biolink:NamedThing"]})
g1.add_node("ENSEMBL:4", **{"category": ["biolink:Gene"]})
g1.add_node("ENSEMBL:6", **{"category": ["biolink:Gene"]})
g1.add_node("HGNC:7", **{"category": ["biolink:Disease"]})
g1.add_node("NCBIGene:8", **{"category": ["biolink:Gene"]})
g1.add_edge(
"ENSEMBL:4",
"HGNC:1",
edge_key=generate_edge_key("ENSEMBL:4", "biolink:same_as", "HGNC:1"),
**{"predicate": "biolink:same_as", "relation": "owl:equivalentClass"}
)
g1.add_edge(
"NCBIGene:3",
"HGNC:1",
edge_key=generate_edge_key("NCBIGene:3", "biolink:same_as", "HGNC:1"),
**{"predicate": "biolink:same_as", "relation": "owl:equivalentClass"}
)
g1.add_edge(
"OMIM:2",
"HGNC:1",
edge_key=generate_edge_key("OMIM:2", "biolink:same_as", "HGNC:1"),
**{"predicate": "biolink:same_as", "relation": "owl:equivalentClass"}
)
g1.add_edge(
"ENSEMBL:6",
"NCBIGene:8",
edge_key=generate_edge_key("ENSEMBL:6", "biolink:same_as", "NCBIGene:8"),
**{"predicate": "biolink:same_as", "relation": "owl:equivalentClass"}
)
g1.add_edge(
"HGNC:7",
"NCBIGene:8",
edge_key=generate_edge_key("HGNC:7", "biolink:same_as", "NCBIGene:8"),
**{"predicate": "biolink:same_as", "relation": "owl:equivalentClass"}
)
updated_graph, clique_graph = clique_merge(
target_graph=g1, prefix_prioritization_map=ppm
)
assert updated_graph.number_of_nodes() == 4
assert updated_graph.number_of_edges() == 2
assert updated_graph.has_node("HGNC:1")
assert updated_graph.has_node("NCBIGene:8")
n1 = updated_graph.nodes()["HGNC:1"]
assert "NCBIGene:3" in n1["same_as"]
assert "ENSEMBL:4" in n1["same_as"]
assert "OMIM:2" not in n1["same_as"]
n2 = updated_graph.nodes()["NCBIGene:8"]
assert "ENSEMBL:6" in n2["same_as"]
assert updated_graph.has_node("OMIM:2")
assert not updated_graph.has_node("NCBIGene:3")
assert not updated_graph.has_node("ENSEMBL:4")
assert updated_graph.has_node("HGNC:7")