def test_generate_set(rosetta):
drug_name = 'test_drug'
drug_name_node = KNode('{}.{}'.format(node_types.DRUG_NAME, drug_name),
node_types.DRUG_NAME)
drug_identifiers = ['CTD:123']
disease_name = 'test_disease'
disease_name_node = KNode(
'{}.{}'.format(node_types.DISEASE_NAME, disease_name),
node_types.DISEASE_NAME)
disease_identifiers = ['DOID:123']
query = UserQuery(drug_identifiers, node_types.DRUG, drug_name_node)
query.add_transition(node_types.GENE)
query.add_transition(node_types.PROCESS)
query.add_transition(node_types.CELL)
query.add_transition(node_types.ANATOMY)
query.add_transition(node_types.PHENOTYPE)
query.add_transition(node_types.DISEASE, end_values=disease_identifiers)
query.add_end_lookup_node(disease_name_node)
d = query.definition
l, r = d.generate_paired_query(4)
assert len(l.transitions) == 4
assert len(r.transitions) == 2
lq = OneSidedLinearUserQuerySet(l)
rq = OneSidedLinearUserQuerySet(r)
#print(lq.generate_cypher()[0])
#print()
#print(rq.generate_cypher()[0])
print(rq.generate_cypher()[0])
assert lq.compile_query(rosetta)
assert rq.compile_query(rosetta)
def test_batch_sequence_variant_to_gene(myvariant):
variant_node = KNode('MYVARIANT_HG38:chr11:g.68032291C>G', type=node_types.SEQUENCE_VARIANT)
variant_node2 = KNode('MYVARIANT_HG38:chrX:g.32389644G>A', type=node_types.SEQUENCE_VARIANT)
variant_node3 = KNode('MYVARIANT_HG38:chr17:g.7674894G>A', type=node_types.SEQUENCE_VARIANT)
batch_annotations = myvariant.batch_sequence_variant_to_gene([variant_node, variant_node2, variant_node3])
relations = batch_annotations['MYVARIANT_HG38:chr11:g.68032291C>G']
identifiers = [node.id for r,node in relations]
assert 'HGNC:7715' in identifiers
predicates = [ relation.standard_predicate for relation,n in relations ]
plabels = [p.label for p in predicates]
assert 'is_missense_variant_of' in plabels
relations = batch_annotations['MYVARIANT_HG38:chrX:g.32389644G>A']
identifiers = [node.id for r,node in relations]
assert 'HGNC:2928' in identifiers
predicates = [ relation.standard_predicate for relation,n in relations ]
plabels = [p.label for p in predicates]
assert 'is_nonsense_variant_of' in plabels
relations = batch_annotations['MYVARIANT_HG38:chr17:g.7674894G>A']
identifiers = [node.id for r,node in relations]
assert 'HGNC:11998' in identifiers
predicates = [ relation.standard_predicate for relation,n in relations ]
plabels = [p.label for p in predicates]
assert 'is_nonsense_variant_of' in plabels
def test_two_sided_query_compose(rosetta):
"""Create a 2sided query, composing it by hand.Mimics what should happen
automatically in a 2 sided user query"""
drug_name = 'test_drug'
drug_name_node = KNode('{}.{}'.format(node_types.DRUG_NAME, drug_name),
node_types.DRUG_NAME)
drug_identifiers = ['CTD:123']
disease_name = 'test_disease'
disease_name_node = KNode(
'{}.{}'.format(node_types.DISEASE_NAME, disease_name),
node_types.DISEASE_NAME)
disease_identifiers = ['DOID:123']
#This will create a OneSidedQuerySet
queryl = UserQuery(drug_identifiers, node_types.DRUG, drug_name_node)
queryl.add_transition(node_types.GENE)
queryl.add_transition(node_types.PROCESS)
queryl.add_transition(node_types.CELL)
queryl.add_transition(node_types.ANATOMY)
#this is another
queryr = UserQuery(disease_identifiers, node_types.DISEASE,
disease_name_node)
queryr.add_transition(node_types.PHENOTYPE)
queryr.add_transition(node_types.ANATOMY)
#The individual queries check out
assert queryl.compile_query(rosetta)
assert queryr.compile_query(rosetta)
#The two sided checks out
twolq = TwoSidedLinearUserQuery(queryl.query, queryr.query)
assert twolq.compile_query(rosetta)
twolqs = TwoSidedLinearUserQuerySet()
twolqs.add_query(twolq, rosetta)
#the two sided set checks out
assert twolqs.compile_query(rosetta)
def test_flush_nodes_changing_node():
# exact same test as non changing node except have to get different synonym map from flush_nodes
node = KNode('MESH:D000096', type=node_types.CHEMICAL_SUBSTANCE)
properties = {'a': 'some prop'}
node.properties = properties
bf = BufferedWriter(rosetta_mock)
bf.write_node(node)
def write_transaction_mock(export_func, nodes, types):
print(types)
# make sure this is the right function
assert export_func == export_node_chunk
# make sure we have out node id in there
assert node.id in nodes
# get the node and see if the properties are preserved
assert nodes[node.id].properties == properties
# see if the types are expected
assert nodes[node.id].export_labels == types == frozenset([
"chemical_substance", "named_thing", "biological_entity",
"molecular_entity"
])
session = Mock()
session.write_transaction = write_transaction_mock
# pass the mock tester to bf and let it rip
bf.flush_nodes(session)
# make sure the synonym map we get here to be used for edge correction is sane
assert 'MESH:D000096' in bf.synonym_map
assert bf.synonym_map['MESH:D000096'] == 'CHEBI:15347'
def test_flush_nodes_non_normilizable():
# exact same test as non changing node except have to get different synonym map from flush_nodes
node = KNode('SOME:curie', type=node_types.CHEMICAL_SUBSTANCE)
properties = {'a': 'some prop'}
node.properties = properties
bf = BufferedWriter(rosetta_mock)
bf.write_node(node)
def write_transaction_mock(export_func, nodes, types):
print(types)
# make sure this is the right function
assert export_func == export_node_chunk
# make sure we have out node id in there
assert node.id in nodes
# get the node and see if the properties are preserved
assert nodes[node.id].properties == properties
# see if the types are expected
assert nodes[node.id].export_labels == []
assert types == frozenset()
session = Mock()
session.write_transaction = write_transaction_mock
# pass the mock tester to bf and let it rip
bf.flush_nodes(session)
# make sure the synonym map we get here to be used for edge correction is sane
assert 'SOME:curie' in bf.synonym_map
assert bf.synonym_map['SOME:curie'] == 'SOME:curie'
def a_test_gwascatalog_variant_to_phenotype(gwascatalog, rosetta):
# turned this off for now because it relies on gwascatalog being precached for CAIDs
#relations = rosetta.cache.get('gwascatalog.sequence_variant_to_disease_or_phenotypic_feature(CAID:CA248392703)')
#identifiers = [node.id for r,node in relations]
#assert 'EFO:0002690' in identifiers
#predicates = [ relation.standard_predicate for relation,n in relations ]
#plabels = set( [p.label for p in predicates] )
#assert 'has_phenotype' in plabels
variant_node = KNode('CAID:CA248392703', type=node_types.SEQUENCE_VARIANT)
relations = gwascatalog.sequence_variant_to_disease_or_phenotypic_feature(variant_node)
identifiers = [node.id for r,node in relations]
assert 'EFO:0002690' in identifiers
predicates = [ relation.standard_predicate for relation,n in relations ]
plabels = set( [p.label for p in predicates] )
assert 'has_phenotype' in plabels
variant_node = KNode('CAID:CA16058750', type=node_types.SEQUENCE_VARIANT)
relations = gwascatalog.sequence_variant_to_disease_or_phenotypic_feature(variant_node)
identifiers = [node.id for r,node in relations]
assert 'EFO:0003898' in identifiers
assert 'EFO:0001359' in identifiers
assert 'ORPHANET:1572' in identifiers
names = [node.name for r,node in relations]
assert 'ankylosing spondylitis' in names
assert 'chronic childhood arthritis' in names
publications = [r.publications for r,node in relations]
assert ['PMID:26301688'] in publications
properties = [r.properties for r,node in relations]
assert properties[0]['pvalue'] == 8.0E-11
variant_node = KNode('DBSNP:rs369602258', type=node_types.SEQUENCE_VARIANT)
results = gwascatalog.sequence_variant_to_disease_or_phenotypic_feature(variant_node)
assert len(results) == 0
def parse_edges(self, provided_by, limit=0):
""" Construct KEdges"""
if not provided_by:
raise RuntimeError(
'Error edge property provided by is not specified')
limit_counter = 0
with open(os.path.join(self.cord_dir, 'edges.txt')) as edges_file:
reader = csv.DictReader(edges_file, delimiter='\t')
for edge_raw in reader:
predicate = LabeledID(identifier='SEMMEDDB:ASSOCIATED_WITH',
label='related_to')
source_node = KNode(edge_raw['Term1'])
target_node = KNode(edge_raw['Term2'])
edge = self.create_edge(source_node=source_node,
target_node=target_node,
input_id=edge_raw['Term1'],
provided_by=provided_by,
predicate=predicate,
publications=[],
properties={
'num_publications':
float(edge_raw['Effective_Pubs']),
'enrichment_p':
float(edge_raw['Enrichment_p'])
})
edge.standard_predicate = predicate
limit_counter += 1
if limit and limit_counter > limit:
break
yield limit_counter - 1, edge
def synonymize_knowledge_graph(knowledge_graph):
id_mappings = {}
if 'nodes' in knowledge_graph:
rosetta = rossetta_setup_default()
nodes = knowledge_graph['nodes']
for node in nodes:
id = ':'.join(re.split(r'\..*:', node['id']))
# try and make nodes with single type
nodes = []
if type(node['type']) == type([]):
nodes = [
KNode(id=id, type=node_type) for node_type in node['type']
]
else:
nodes = [KNode(id=id, type=node['type'])]
id_picks = [id]
for n in nodes:
rosetta.synonymizer.synonymize(n)
# if node Id after synonymization is d/t from kg node track change to use that
if n.id not in id_picks:
id_picks.append(n.id)
if 'equivalent_identifiers' not in node:
node['equivalent_identifiers'] = []
node['equivalent_identifiers'].extend([
x[0] for x in list(n.synonyms)
if x[0] not in node['equivalent_identifiers']
])
id_last_change = id_picks[len(id_picks) - 1]
id_mappings[node['id']] = id_last_change
node['id'] = id_last_change
else:
logger.warning('Unable to locate nodes in knowledge graph')
return knowledge_graph, id_mappings
def test_list_returns_zero(omnicorpus):
disease_node = KNode('UBERON:0013694', type=node_types.ANATOMICAL_ENTITY)
go_node = KNode('GO:0045892', type=node_types.BIOLOGICAL_PROCESS)
nodes = [disease_node, go_node]
results = omnicorpus.get_all_shared_pmids(nodes)
assert len(results) == 1
assert len(list(results.values())[0]) == 0
def get_edges_from_file(self, file_name, provided_by, delimiter):
"""
All is stuff is till we get kgx to merge edges. For now creating
a pattern looking like a robokopservice and let writer handle it.
:param file_name:
:return:
"""
if not file_name:
return
bl_resolver = BL_lookup()
with open(file_name) as edge_file:
reader = csv.DictReader(edge_file, delimiter=delimiter)
for raw_edge in reader:
edge_label = raw_edge['edge_label'].split(':')[-1]
relation_predicate = raw_edge['relation']
predicate = LabeledID(
identifier=
relation_predicate, #bl_resolver.resolve_curie(edge_label),
label=edge_label)
source_node = KNode(raw_edge['subject'])
target_node = KNode(raw_edge['object'])
edge = self.create_edge(
source_node=source_node,
target_node=target_node,
input_id=source_node.id,
provided_by=provided_by,
predicate=predicate,
)
edge.standard_predicate = predicate
yield edge
def process_variant_to_gene_relationships(self, variant_nodes: list, writer: WriterDelegator):
all_results = self.genetics_services.get_variant_to_gene(self.crawl_for_service, variant_nodes)
for source_node_id, results in all_results.items():
# convert the simple edges and nodes to rags objects and write them to the graph
for (edge, node) in results:
gene_node = KNode(node.id, type=node.type, name=node.name, properties=node.properties)
if self.recreate_sv_node:
variant_node = KNode(source_node_id, type= node_types.SEQUENCE_VARIANT)
variant_node.add_export_labels([node_types.SEQUENCE_VARIANT])
writer.write_node(variant_node)
if gene_node.id not in self.written_genes:
writer.write_node(gene_node)
self.written_genes.add(gene_node.id)
predicate = LabeledID(identifier=edge.predicate_id, label=edge.predicate_label)
gene_edge = KEdge(source_id=source_node_id,
target_id=gene_node.id,
provided_by=edge.provided_by,
ctime=edge.ctime,
original_predicate=predicate,
# standard_predicate=predicate,
input_id=edge.input_id,
properties=edge.properties)
writer.write_edge(gene_edge)
logger.info(f'added {len(results)} variant relationships for {source_node_id}')
def test_get_biological_process_by_gene_family(panther):
top_family_node = KNode('PTHR11003',
type=node_types.GENE_FAMILY,
name='POTASSIUM CHANNEL, SUBFAMILY K')
sub_family_node = KNode('PTHR11003:SF241',
type=node_types.GENE_FAMILY,
name='POTASSIUM CHANNEL, SUBFAMILY K Member 5')
response = panther.get_biological_process_or_activity_by_gene_family(
top_family_node)
node_ids = [relation[1].id for relation in response]
for edge, node in response:
assert node.type == node_types.BIOLOGICAL_PROCESS_OR_ACTIVITY
#molecular activity
assert 'GO:0005261' in node_ids
#biological process
assert 'GO:0006811' in node_ids
response = panther.get_biological_process_or_activity_by_gene_family(
sub_family_node)
node_ids = [relation[1].id for relation in response]
for edge, node in response:
assert node.type == node_types.BIOLOGICAL_PROCESS_OR_ACTIVITY
#molecular activity
assert 'GO:0005261' in node_ids
#biological process
assert 'GO:0006811' in node_ids
def process_associations(self,
r,
predicate,
target_node_type,
reverse=False):
"""Given a response from biolink, create our edge and node structures.
Sometimes (as in pathway->Genes) biolink returns the query as the object, rather
than the subject. reverse=True will handle this case, bringing back the subject
of the response, rather than the object."""
edge_nodes = []
for association in r['associations']:
pubs = []
if 'publications' in association and association[
'publications'] is not None:
for pub in association['publications']:
# Sometimes, we get back something like "uniprotkb" instead of a PMID. We don't want it.
pubid_prefix = pub['id'][:4].upper()
if pubid_prefix == 'PMID':
pubs.append(pub['id'])
if reverse:
obj = KNode(association['subject']['id'], target_node_type,
association['subject']['label'])
else:
obj = KNode(association['object']['id'], target_node_type,
association['object']['label'])
rel = {
'typeid': association['relation']['id'],
'label': association['relation']['label']
}
props = {'publications': pubs, 'relation': rel}
edge = KEdge('biolink', predicate, props)
edge_nodes.append((edge, obj))
return edge_nodes
def parse_nodes(self, limit=0):
"""
Parse nodes.
:param limit: for testing reads first n nodes from file
:return: dict with node_id as key and KNode as value
"""
print('parsing nodes...')
limit_counter = 0
with open(os.path.join(self.cord_dir, 'nodes.txt')) as nodes_file:
reader = csv.DictReader(nodes_file, delimiter='\t')
for raw_node in reader:
# transform headers to knode attrbutes
labels = raw_node.get('semantic_type')
labels = labels.replace(']', '').replace('[', '').replace(
'\\', '').replace("'", '')
labels = labels.split(',')
node = KNode({
'id': raw_node.get('normalized_curie'),
'type': labels[0],
'name': raw_node.get('name'),
'properties': {
'input_term': raw_node.get('input_term')
}
})
node.add_export_labels(labels)
limit_counter += 1
if limit and limit_counter > limit:
break
yield limit_counter - 1, node
def test_phenotype(rosetta):
node = KNode("MEDDRA:10014408", type=node_types.PHENOTYPIC_FEATURE)
synonymize(node, rosetta.core)
assert len(node.synonyms) > 10
hpsyns = node.get_synonyms_by_prefix("HP")
assert len(hpsyns) > 0
print(hpsyns)
def test_hgnc_label(rosetta):
"""Do I get a label back?"""
node = KNode('HGNC:18729', type=node_types.GENE)
rosetta.synonymizer.synonymize(node)
hgnc = node.get_synonyms_by_prefix('HGNC')
assert node.name is not None
assert node.name != ''
def term_get_ancestors(self, node_type, root_iri):
results = self.triplestore.query_template(
template_text=self.query,
inputs={'root_uri': root_iri},
outputs=['parent_id', 'parent_label', 'child_id', 'child_label'])
print('found total ', len(results), ' results.')
nodes = set()
edges = set()
for index, row in enumerate(results):
# Output type would be same as input type?
ancestor_node = KNode(Text.obo_to_curie(row['parent_id']),
name=row['parent_label'],
type=node_type)
child_node = KNode(Text.obo_to_curie(row['child_id']),
name=row['child_label'],
type=node_type)
if ancestor_node.id == child_node.id:
# refrain from adding edge to the node itself
continue
predicate = LabeledID(identifier='rdfs:subClassOf',
label='subclass of')
edge = self.create_edge(
source_node=child_node,
target_node=ancestor_node,
predicate=predicate,
provided_by='uberongraph.term_get_ancestors',
input_id=child_node.id)
nodes.add(child_node)
nodes.add(ancestor_node)
edges.add(edge)
return nodes, edges
def create_node(session):
#Make sure we're clean
session.run("MATCH (a {id:{id}}) DETACH DELETE a", {"id": TEST_ID})
original = get_node(TEST_ID, session)
assert original is None
node = KNode(TEST_ID, node_types.DISEASE)
node.add_synonyms(ORIGINAL_SYNONYMS)
export_node(node, session)
def x_test_event_to_drug(mychem):
node = KNode('HP:0002018',
type=node_types.PHENOTYPIC_FEATURE,
name='Nausea')
node.add_synonyms(
set([LabeledID(identifier='MedDRA:10028813', label='Nausea')]))
results = mychem.get_drug_from_adverse_events(node)
assert len(results) > 0
def x_test_event_to_drug(mychem):
node = KNode('MONDO:0002050',
type=node_types.DISEASE,
name='Mental Depression')
node.add_synonyms(
set([LabeledID(identifier='MedDRA:10002855', label='Depression')]))
results = mychem.get_drug_from_adverse_events(node)
assert len(results) > 0
def test_list_with_bad_curie(omnicorpus):
node = KNode('CL:0000097', type=node_types.CELL)
disease_node = KNode('MONDO:0004979', type=node_types.DISEASE)
drug_node = KNode('CHEBI:45783', type=node_types.CHEMICAL_SUBSTANCE)
stinker = KNode('FAKEO:102830', type=node_types.CHEMICAL_SUBSTANCE)
nodes = [node, disease_node, drug_node, stinker]
results = omnicorpus.get_all_shared_pmids(nodes)
assert len(results) == 6
def future_test_disease_normalization(rosetta):
node = KNode('DOID:4325', type=node_types.DISEASE)
synonyms = synonymize(node, rosetta.core)
print(synonyms)
node.add_synonyms(synonyms)
mondos = node.get_synonyms_by_prefix('MONDO')
assert len(mondos) > 0
assert Text.get_curie(node.id) == 'MONDO'
def test2():
from greent.rosetta import Rosetta
rosetta = Rosetta()
gt = rosetta.core
support = ChemotextSupport(gt)
from greent.graph_components import KNode
node_a = KNode('CTD:1,2-linoleoylphosphatidylcholine', type=node_types.CHEMICAL_SUBSTANCE, name='1,2-linoleoylphosphatidylcholine')
node_b = KNode('CTD:Hydrogen Peroxide', type=node_types.CHEMICAL_SUBSTANCE, name='Hydrogen Peroxide')
def xxtest_go(rosetta):
node = KNode("HGNC:10593",label="SCN5A",type=node_types.GENE)
s3 = rosetta.cache.get('synonymize(HGNC:10593)')
rosetta.synonymizer.synonymize(node)
print (node.get_synonyms_by_prefix('UNIPROTKB'))
biolink = rosetta.core.biolink
r=biolink.gene_get_process_or_function(node)
assert len(r) > 0
def test_gene_to_drug_synonym(ctd):
# Even though the main identifier is drugbank, CTD should find the right synonym in there somewhere.
input_node = KNode("DB:FakeID", type=node_types.GENE)
input_node.add_synonyms(set(["NCBIGene:5743"]))
results = ctd.gene_to_drug(input_node)
for _, node in results:
assert node.type == node_types.DRUG
result_ids = [node.id for edge, node in results]
assert 'MESH:D000068579' in result_ids # Cox2 for a cox2 inhibitor
def test_complicated(rosetta):
"""make sure that a very complicated cast gets everything to the right place"""
fname = 'caster.output_filter(input_filter(upcast(hetio~disease_to_phenotype,disease_or_phenotypic_feature),disease,typecheck~is_disease),disease,typecheck~is_disease)'
func = rosetta.get_ops(fname)
assert func is not None
node = KNode('HP:0007354', type=node_types.PHENOTYPIC_FEATURE)
node.add_synonyms(set([LabeledID(identifier='DOID:332', label='ALS')]))
results = func(node)
assert results is not None
def test_smdb_id_normalizer(hmdb):
node = KNode('HMDB:HMDB0112245', type=node_types.CHEMICAL_SUBSTANCE)
node.add_synonyms(['HMDB:HMDB0112245'])
results = hmdb.metabolite_to_pathway(node)
for r in results:
r_node = r[1]
if r_node.id.startswith('SMPDB'):
uncuried = Text.un_curie(r_node.id)
assert len(uncuried) == 10
def test_uniprot(rosetta):
"""Do we correctly synonymize if all we have is a UniProtKB identifier?"""
node = KNode('UniProtKB:O75381', type=node_types.GENE)
rosetta.synonymizer.synonymize(node)
hgnc = node.get_synonyms_by_prefix('HGNC')
assert len(hgnc) == 1
assert hgnc.pop() == 'HGNC:8856'
assert node.id == 'HGNC:8856'
assert node.name == 'PEX14'
def test_crappy_uniprot(rosetta):
"""Do we correctly synonymize if all we have is a UniProtKB identifier?"""
node = KNode('UniProtKB:A0A024QZH5', type=node_types.GENE)
rosetta.synonymizer.synonymize(node)
hgnc = node.get_synonyms_by_prefix('HGNC')
assert len(hgnc) == 1
assert hgnc.pop() == 'HGNC:18859'
assert node.id == 'HGNC:18859'
assert node.name == 'SPHK2'
def test_failing_uniprot_2(rosetta):
"""Do we correctly synonymize if all we have is a UniProtKB identifier?"""
node = KNode('UniProtKB:P14416', type=node_types.GENE, name='')
rosetta.synonymizer.synonymize(node)
hgnc = node.get_synonyms_by_prefix('HGNC')
assert len(hgnc) == 1
assert hgnc.pop() == 'HGNC:3023'
assert node.id == 'HGNC:3023'
assert node.name == 'DRD2'
