def __init__(self, context):
super(OntologicalHeirarchy, self).__init__("ontological_hierarchy",
context)
self.triplestore = TripleStore(self.url)
self.prefix_set = {
node_types.DISEASE_OR_PHENOTYPIC_FEATURE: ['HP', 'MONDO'],
node_types.CELLULAR_COMPONENT: ['CL'],
node_types.BIOLOGICAL_PROCESS_OR_ACTIVITY: ['GO'],
node_types.ANATOMICAL_ENTITY: ['UBERON'],
node_types.CHEMICAL_SUBSTANCE: ['CHEBI']
}
self.root_uris = {
node_types.ANATOMICAL_ENTITY:
"<http://purl.obolibrary.org/obo/UBERON_0001062>",
node_types.DISEASE:
"<http://purl.obolibrary.org/obo/MONDO_0000001>",
node_types.MOLECULAR_ACTIVITY:
"<http://purl.obolibrary.org/obo/GO_0003674>",
node_types.BIOLOGICAL_PROCESS:
"<http://purl.obolibrary.org/obo/GO_0008150>",
node_types.CHEMICAL_SUBSTANCE:
"<http://purl.obolibrary.org/obo/CHEBI_24431>",
node_types.PHENOTYPIC_FEATURE:
"<http://purl.obolibrary.org/obo/HP_0000118>",
node_types.CELL:
"http://purl.obolibrary.org/obo/CL_0000000",
node_types.CELLULAR_COMPONENT:
"http://purl.orolibrary.org/obo/GO_0005575"
}
obo_prefixes = '\n'.join([
f'PREFIX {pref}: <http://purl.obolibrary.org/obo/{pref}_>'
for pref in set(
reduce(lambda x, y: x + y, self.prefix_set.values(), []))
])
self.query = f"""
def __init__(self, context, obo):
""" Load an obo file. """
super(GenericOntology, self).__init__("go", context)
self.url = context.config.get_service('uberongraph').get("url", None)
self.triplestore = TripleStore(self.url)
self.sparql_wrapper = SPARQLWrapper(self.url)
curie_prefix_map = Curie_Resolver.get_curie_to_uri_map()
self.resolve_uri = Curie_Resolver.uri_to_curie
self.ontology_prefixes = list(
map(lambda x: f'PREFIX {x}: <{curie_prefix_map[x]}>',
curie_prefix_map))
def __init__(self, context): #triplestore):
super(UberonGraphKS, self).__init__("uberongraph", context)
self.triplestore = TripleStore(self.url)
#TODO: Pull this from the biolink model?
self.class_defs = {
node_types.CELL: 'CL:0000000',
node_types.ANATOMICAL_ENTITY: 'UBERON:0001062',
node_types.BIOLOGICAL_PROCESS: 'GO:0008150',
node_types.MOLECULAR_ACTIVITY: 'GO:0003674',
node_types.CHEMICAL_SUBSTANCE: 'CHEBI:24431',
node_types.DISEASE: 'MONDO:0000001',
node_types.PHENOTYPIC_FEATURE: 'UPHENO:0001002'
}
def __init__(self):
self.url = "https://stars-app.renci.org/uberongraph/sparql"
self.triplestore = TripleStore(self.url)
self.prefix_set = {
node_types.DISEASE_OR_PHENOTYPIC_FEATURE: ['HP', 'MONDO'],
node_types.CELLULAR_COMPONENT: ['CL'],
node_types.BIOLOGICAL_PROCESS_OR_ACTIVITY: ['GO'],
node_types.ANATOMICAL_ENTITY: ['UBERON'],
node_types.CHEMICAL_SUBSTANCE: ['CHEBI']
}
self.root_uris = {
node_types.ANATOMICAL_ENTITY:
"<http://purl.obolibrary.org/obo/UBERON_0001062>",
node_types.DISEASE:
"<http://purl.obolibrary.org/obo/MONDO_0000001>",
node_types.MOLECULAR_ACTIVITY:
"<http://purl.obolibrary.org/obo/GO_0003674>",
node_types.BIOLOGICAL_PROCESS:
"<http://purl.obolibrary.org/obo/GO_0008150>",
node_types.CHEMICAL_SUBSTANCE:
"<http://purl.obolibrary.org/obo/CHEBI_24431>",
node_types.PHENOTYPIC_FEATURE:
"<http://purl.obolibrary.org/obo/HP_0000118>",
node_types.CELL:
"<http://purl.obolibrary.org/obo/CL_0000000>",
node_types.CELLULAR_COMPONENT:
"<http://purl.orolibrary.org/obo/GO_0005575>"
}
obo_prefixes = '\n'.join([
f'PREFIX {pref}: <http://purl.obolibrary.org/obo/{pref}_>'
for pref in set(
reduce(lambda x, y: x + y, self.prefix_set.values(), []))
])
self.query = f"""
{obo_prefixes}
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
select ?parent_id ?parent_label ?child_id ?child_label
where {{
?parent_id rdfs:subClassOf $root_uri .
?child_id rdfs:subClassOf ?parent_id.
OPTIONAL {{
?parent_id rdfs:label ?parent_label.
?child_id rdfs:label ?child_label.
}}
}}
"""
rosetta = Rosetta()
self.wdg = WriterDelegator(rosetta)
class MeshKS(Service):
def __init__(self, context, url):
super(MeshKS, self).__init__("mesh", context)
self.triplestore = TripleStore(self.url)
def get_label_by_id(self, term_id):
result = self.triplestore.query_template(inputs={"term_id": term_id},
outputs=['label'],
template_text="""
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
PREFIX owl: <http://www.w3.org/2002/07/owl#>
PREFIX meshv: <http://id.nlm.nih.gov/mesh/vocab#>
PREFIX mesh: <http://id.nlm.nih.gov/mesh/>
PREFIX mesh2015: <http://id.nlm.nih.gov/mesh/2015/>
PREFIX mesh2016: <http://id.nlm.nih.gov/mesh/2016/>
PREFIX mesh2017: <http://id.nlm.nih.gov/mesh/2017/>
PREFIX mesh2018: <http://id.nlm.nih.gov/mesh/2018/>
SELECT DISTINCT ?label
FROM <http://id.nlm.nih.gov/mesh>
WHERE {
VALUES (?id) { ( $term_id ) }
?id rdfs:label ?label .
}
ORDER BY ?label
""")
return list(map(lambda r: {'label': r['label']}, result))
class MeSH(object):
def __init__(self, uri="http://id.nlm.nih.gov/mesh/sparql"):
self.triplestore = TripleStore (uri)
def get_broader (self, term):
return self.triplestore.query_template (
inputs={ "term" : term, "prefixes" : self.get_prefixes () },
outputs= [ "obj", "name" ],
template_text="""
$prefixes
SELECT DISTINCT ?obj ?name ?itemName FROM <http://id.nlm.nih.gov/mesh>
WHERE {
?item meshv:broaderDescriptor ?obj ;
rdfs:label ?itemName.
?obj rdfs:label ?name .
filter (regex(lcase(str(?itemName)), lcase(str("$term"))))
}
ORDER BY ?p
""")
"""
SELECT DISTINCT ?obj ?name FROM <http://id.nlm.nih.gov/mesh>
WHERE {
$term meshv:broaderDescriptor ?obj .
?obj rdfs:label ?name .
}
ORDER BY ?p
"""
def get_prefixes (self):
return """
def __init__(self, context): #triplestore):
super(OmniCorp, self).__init__("omnicorp", context)
self.triplestore = TripleStore(self.url)
self.prefix_to_uri = {
'UBERON': 'http://purl.obolibrary.org/obo/UBERON_',
'BSPO': 'http://purl.obolibrary.org/obo/BSPO_',
'PATO': 'http://purl.obolibrary.org/obo/PATO_',
'GO': 'http://purl.obolibrary.org/obo/GO_',
'MONDO': 'http://purl.obolibrary.org/obo/MONDO_',
'HP': 'http://purl.obolibrary.org/obo/HP_',
'ENVO:': 'http://purl.obolibrary.org/obo/ENVO_',
'OBI': 'http://purl.obolibrary.org/obo/OBI_',
'CL': 'http://purl.obolibrary.org/obo/CL_',
'SO': 'http://purl.obolibrary.org/obo/SO_',
'CHEBI': 'http://purl.obolibrary.org/obo/CHEBI_',
'HGNC': 'http://identifiers.org/hgnc/HGNC:',
'MESH': 'http://id.nlm.nih.gov/mesh/'
}
class GenericOntology(Service):
""" Sure, don't just dig around in obo files they say. But when the SPARQL is dry, we will drink straight from the obo if need be. """
def __init__(self, context, obo):
""" Load an obo file. """
super(GenericOntology, self).__init__("go", context)
self.url = context.config.get_service('uberongraph').get("url", None)
self.triplestore = TripleStore(self.url)
self.sparql_wrapper = SPARQLWrapper(self.url)
curie_prefix_map = Curie_Resolver.get_curie_to_uri_map()
self.resolve_uri = Curie_Resolver.uri_to_curie
self.ontology_prefixes = list(
map(lambda x: f'PREFIX {x}: <{curie_prefix_map[x]}>',
curie_prefix_map))
def add_sparql_prefixes(self, query_template):
return '\n'.join(self.ontology_prefixes) + '\n' + query_template
def run_sparql_query_raw(self, query):
query = self.add_sparql_prefixes(query)
self.sparql_wrapper.setQuery(query)
self.sparql_wrapper.setReturnFormat(JSON)
results = self.sparql_wrapper.query().convert()
return results
def query_sparql(self, query_template, inputs, outputs):
# prepend prefixes here to avoid every one doing the same thing
q = self.add_sparql_prefixes(query_template)
logger.error(q)
return self.triplestore.query_template(template_text=q,
inputs=inputs,
outputs=outputs)
def label(self, identifier):
"""Return the label for an identifier"""
query_text = f"""
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT DISTINCT ?labels
FROM <http://reasoner.renci.org/ontology>
WHERE {{
$identifier rdfs:label ?labels.
}}
"""
results = self.query_sparql(query_text,
inputs={'identifier': identifier},
outputs=['labels'])
return results[0]["labels"] if len(results) > 0 else ''
def is_a(self, identifier, ancestors):
"""Determine whether a term has a particular ancestor"""
query_template = lambda ancestor: f"""
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
ASK
FROM <http://reasoner.renci.org/ontology/closure>
FROM <http://reasoner.renci.org/ontology>
WHERE {{
{identifier} rdfs:subClassOf {ancestor}.
}}
"""
result = []
for ancestor in ancestors.split(','):
ancestor = ancestor.strip(' ')
response = self.run_sparql_query_raw(query_template(ancestor))
if response['boolean']:
result.append(ancestor)
is_a = len(result) > 0
return is_a, result
def single_level_is_a(self, identifier):
""" Get single-level 'is_a' descendants. """
query_text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX ID: <http://www.geneontology.org/formats/oboInOwl#id>
SELECT DISTINCT ?descendant ?descendant_id
FROM <http://reasoner.renci.org/ontology>
WHERE {
?descendant rdfs:subClassOf $identifier
OPTIONAL {
?descendant ID: ?descendant_id
}
}
"""
results = self.query_sparql(query_text,
inputs={'identifier': identifier},
outputs=['descendant', 'descendant_id'])
response_curies = reduce(
lambda x, y: x + [
y['descendant_id']
if 'descendant_id' in y else self.resolve_uri(y['descendant'])
], results, [])
return response_curies
def descendants(self, identifier):
""" This is also known as a recursive-'is_a' function, returning all levels below the input"""
query_text = f"""
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX ID: <http://www.geneontology.org/formats/oboInOwl#id>
SELECT DISTINCT ?descendant ?descendant_id
FROM <http://reasoner.renci.org/ontology/closure>
WHERE {{
?descendant rdfs:subClassOf $identifier.
OPTIONAL {{ ?descendant ID: ?descendant_id. }}
}}
"""
results = self.query_sparql(query_template=query_text,
inputs={'identifier': identifier},
outputs=['descendant', 'descendant_id'])
result_list = reduce(
lambda x, y: x + [
y['descendant_id']
if 'descendant_id' in y else self.resolve_uri(y['descendant'])
], results, [])
return result_list
def ancestors(self, identifier):
""" This is also known as a recursive-'is_a' function, returning all levels below the input"""
query_text = f"""
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX ID: <http://www.geneontology.org/formats/oboInOwl#id>
SELECT DISTINCT ?ancestor ?ancestor_id
FROM <http://reasoner.renci.org/ontology/closure>
WHERE {{
$identifier rdfs:subClassOf ?ancestor.
OPTIONAL {{
?ancestor ID: ?ancestor_id.
}}
}}
"""
results = self.query_sparql(query_template=query_text,
inputs={'identifier': identifier},
outputs=['ancestor', 'ancestor_id'])
result_list = reduce(
lambda x, y: x + [
y['ancestor_id']
if 'ancestor_id' in y else self.resolve_uri(y['ancestor'])
], results, [])
return result_list
def xrefs(self, identifier):
""" Get external references. """
query_text = f"""prefix xref: <http://www.geneontology.org/formats/oboInOwl#hasDbXref>
SELECT DISTINCT ?xrefs
FROM <http://reasoner.renci.org/ontology>
WHERE {{
$identifier xref: ?xrefs
}}
"""
results = self.query_sparql(query_template=query_text,
inputs={'identifier': identifier},
outputs=['xrefs'])
results = reduce(lambda x, y: x + [y['xrefs']], results, [])
return results
def synonyms(self, identifier, curie_pattern=None):
""" Get synonyms. """
query_template = lambda predicate: f"""
PREFIX RELATED_SYNONYM: <http://www.geneontology.org/formats/oboInOwl#hasRelatedSynonym>
PREFIX EXACT_SYNONYM: <http://www.geneontology.org/formats/oboInOwl#hasExactSynonym>
PREFIX XREF: <http://www.geneontology.org/formats/oboInOwl#hasDbXref>
PREFIX DEFENITION: <http://purl.obolibrary.org/obo/IAO_0000115>
SELECT DISTINCT ?desc ?xref ?defn
WHERE {{
$identifier {predicate} ?desc.
OPTIONAL {{
?desc XREF: ?xref.
?desc DEFENITION: ?defn.
}}
}}
"""
exact = self.query_sparql(
query_template=query_template("EXACT_SYNONYM:"),
inputs={'identifier': identifier},
outputs=['desc', 'xref', 'defn'])
for row in exact:
row['scope'] = 'EXACT'
related = self.query_sparql(
query_template=query_template("RELATED_SYNONYM:"),
inputs={'identifier': identifier},
outputs=['desc', 'xref', 'defn'])
for row in related:
row['scope'] = 'RELATED'
return exact + related
def search(self, text, is_regex=False, ignore_case=True):
""" Search for the text, treating it as a regular expression if indicated. """
search_string = text
if is_regex and ignore_case:
filtr = f"""
(
regex(str(?definition), "$search_string","i") ||
regex(str(?label), "$search_string","i") ||
regex(str(?related_synonym), "$search_string","i") ||
regex(str(?exact_synonym), "$search_string","i")
)"""
elif is_regex and not ignore_case:
filtr = f"""
(
regex(str(?definition), "$search_string") ||
regex(str(?label), "$search_string") ||
regex(str(?related_synonym), "$search_string") ||
regex(str(?exact_synonym), "$search_string")
)
"""
elif not is_regex and ignore_case:
search_string = search_string.lower()
filtr = f"""
(
lcase(str(?label))= "$search_string" ||
lcase(str(?definition))= "$search_string" ||
lcase(str(?related_synonym))= "$search_string" ||
lcase(str(?exact_synonym))= "$search_string"
)
"""
else:
filtr = f"""
(
str(?label) = "$search_string" ||
str(?definition) = "$search_string" ||
str(?exact_synonym) = "$search_string" ||
str(?related_synonym) = "$search_string"
)"""
query_text = f"""
PREFIX DEFINED_BY: <http://www.w3.org/2000/01/rdf-schema#isDefinedBy>
PREFIX DEFINITION: <http://purl.obolibrary.org/obo/IAO_0000115>
PREFIX RELATED_SYNONYM: <http://www.geneontology.org/formats/oboInOwl#hasRelatedSynonym>
PREFIX EXACT_SYNONYM: <http://www.geneontology.org/formats/oboInOwl#hasExactSynonym>
SELECT DISTINCT ?id ?label ?definition ?defined_by
WHERE {{
OPTIONAL{{
?id EXACT_SYNONYM: ?exact_synonym.
}}
OPTIONAL {{
?id RELATED_SYNONYM: ?related_synonym.
}}
OPTIONAL {{
?id rdfs:label ?label.
}}
OPTIONAL {{
?id DEFINITION: ?definition.
}}
OPTIONAL {{
?id DEFINED_BY: ?defined_by.
}}
FILTER {filtr}.
}}
"""
response = self.query_sparql(
query_template=query_text,
inputs={'search_string': search_string},
outputs=['id', 'label', 'defined_by', 'definition'])
for row in response:
row['id'] = Curie_Resolver.uri_to_curie(row['id'])
return response
def lookup(self, identifier):
""" Given an identifier, find ids in the ontology for which it is an xref. """
assert identifier and ':' in identifier, "Must provide a valid curie. Curie must have format " \
"<PREFIX>:<ID>"
query_template = """
PREFIX XREF: <http://www.geneontology.org/formats/oboInOwl#hasDbXref>
PREFIX LABEL: <http://www.w3.org/2000/01/rdf-schema#label>
PREFIX ID: <http://www.geneontology.org/formats/oboInOwl#id>
SELECT DISTINCT ?xrefs ?term_id ?term_label
FROM <http://reasoner.renci.org/ontology>
WHERE {
?term XREF: ?o;
XREF: ?xrefs;
ID: ?term_id;
LABEL: ?term_label.
FILTER(?o = '$identifier' && !isBlank(?term)).
}
"""
result = self.query_sparql(query_template=query_template,
inputs={'identifier': identifier},
outputs=['xrefs', 'term_id', 'term_label'])
response = []
buffer = {}
for row in result:
if row['term_id'] not in buffer:
buffer[row['term_id']] = {
'label': row['term_label'],
'xrefs': []
}
xref = row['xrefs']
buffer[row['term_id']]['xrefs'] += [
row['xrefs']
] if row['xrefs'] not in buffer[row['term_id']]['xrefs'] else []
for term_id in buffer:
response.append({
'id': term_id,
'label': buffer[term_id]['label'],
'xrefs': buffer[term_id]['xrefs']
})
return response
def id_list(self, identifier):
identifier_uri = Curie_Resolver.get_curie_to_uri_map().get(
identifier.upper(), None)
if identifier_uri == None:
return []
query = f"""
PREFIX TYPE: <http://www.w3.org/1999/02/22-rdf-syntax-ns#type>
PREFIX CLASS: <http://www.w3.org/2002/07/owl#Class>
PREFIX ID: <http://www.geneontology.org/formats/oboInOwl#id>
SELECT DISTINCT ?term ?term_id
FROM <http://reasoner.renci.org/ontology>
WHERE {{
?term TYPE: CLASS:.
FILTER (strstarts(lcase(str(?term)), lcase("$identifier")))
OPTIONAL {{
?term ID: ?term_id #try to get the id from sparql else parse ??
}}
}}
"""
result = self.query_sparql(query_template=query,
inputs={'identifier': identifier_uri},
outputs=['term', 'term_id'])
return reduce(
lambda x, y: x +
[y['term_id']
if 'term_id' in y else self.resolve_uri(y['term'])], result, [])
def exactMatch(self, identifier):
#if ontolgies are missing we should be looking here
query_string = lambda predicate: f"""
PREFIX EXACT_MATCH: <http://www.w3.org/2004/02/skos/core#exactMatch>
PREFIX EQUIVALENT_CLASS: <http://www.w3.org/2002/07/owl#equivalentClass>
PREFIX ID: <http://www.geneontology.org/formats/oboInOwl#id>
SELECT DISTINCT ?match ?match_id
FROM <http://reasoner.renci.org/ontology>
WHERE {{
$identifier {predicate} ?match.
OPTIONAL {{
?match ID: ?match_id.
}}
FILTER (!isBlank(?match)) #This sometimes returns blank nodes
}}
"""
result = reduce(
lambda x, y: x + [
y['match_id']
if 'match_id' in y else self.resolve_uri(y['match'])
],
self.query_sparql(query_template=query_string('EXACT_MATCH:'),
inputs={'identifier': identifier},
outputs=['match', 'match_id']), [])
result += list(
filter(
lambda x: x not in result,
reduce(
lambda x, y: x + [
y['match_id']
if 'match_id' in y else self.resolve_uri(y['match'])
],
self.query_sparql(
query_template=query_string('EQUIVALENT_CLASS:'),
inputs={'identifier': identifier},
outputs=['match', 'match_id']), [])))
return result
def closeMatch(self, identifier):
query_template = """
PREFIX CLOSE_MATCH: <http://www.w3.org/2004/02/skos/core#closeMatch>
PREFIX ID: <http://www.geneontology.org/formats/oboInOwl#id>
SELECT DISTINCT ?match ?match_id
FROM <http://reasoner.renci.org/ontology>
WHERE {
$identifier CLOSE_MATCH: ?match.
OPTIONAL {
?match ID: ?match_id.
}
FILTER (!isBlank(?match)) #This sometimes returns blank nodes
}
"""
results = reduce(
lambda x, y: x + [
y['match_id']
if 'match_id' in y else self.resolve_uri(y['match'])
],
self.query_sparql(query_template=query_template,
inputs={'identifier': identifier},
outputs=['match', 'match_id']), [])
return results
def subterms(self, identifier):
return self.descendants(identifier)
def superterms(self, identifier):
return self.ancestors(identifier)
def parents(self, identifier):
"""First generation ancestors"""
query_template = f"""PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX ID: <http://www.geneontology.org/formats/oboInOwl#id>
SELECT DISTINCT ?parent ?parent_id
FROM <http://reasoner.renci.org/ontology>
WHERE {{
$identifier rdfs:subClassOf ?parent.
OPTIONAL {{
?parent ID: ?parent_id
}}
FILTER(!isBlank(?parent))
}}"""
result = reduce(
lambda x, y: x + [
y['parent_id']
if 'parent_id' in y else self.resolve_uri(y['parent'])
],
self.query_sparql(query_template=query_template,
inputs={'identifier': identifier},
outputs=['parent', 'parent_id']), [])
return result
def children(self, identifier):
"""first generation descedants"""
result = self.single_level_is_a(identifier)
return result
def siblings(self, identifier):
"""
Common parents
"""
parents = self.parents(identifier)
sibilings = []
for parent in parents:
sibilings += list(
filter(
lambda x: x != identifier and x not in sibilings,
self.children(parent if 'http' not in
parent else f'<{parent}>')))
return sibilings
def property_value(self, identifier, property_key):
""" Get properties """
query_template = """
SELECT ?property_value
WHERE {
$identifier <$property_key> ?property_value.
}
"""
result = self.query_sparql(query_template=query_template,
inputs={
'identifier': identifier,
'property_key': property_key
},
outputs=['property_value'])
response = reduce(lambda x, y: x + [y['property_value']], result, [])
if len(response) == 1:
return response[0]
else:
return response
def all_properties(self, identifier):
""" Get ALL properties for a CURIE """
query_template = """
SELECT ?property_key ?property_value ?property_label
FROM <http://reasoner.renci.org/ontology>
WHERE
{
$identifier ?property_key ?property_value.
OPTIONAL {
?property_key rdfs:label ?property_label.
}
FILTER (!isBlank(?property_value))
}
"""
results = self.query_sparql(
query_template=query_template,
inputs={'identifier': identifier},
outputs={'property_value', 'property_key', 'property_label'})
# group it by property label for those which have label
grouped = {}
for row in results:
label = row['property_label'] if 'property_label' in row else None
key = row['property_key']
if key not in grouped:
grouped[key] = {'property_label': label, 'property_values': []}
if row['property_value'] not in grouped[key]['property_values']:
grouped[key]['property_values'].append(row['property_value'])
for key in grouped:
grouped[key].update({'property_key': key})
return list(map(lambda x: grouped[x], grouped))
def __init__(self, context): #triplestore):
super(ChemBioKS, self).__init__("chembio", context)
self.triplestore = TripleStore(self.url)
class ChemBioKS(Service):
""" Generic service endpoints for medical and bio-chemical data. This set
comprises portions of chem2bio2rdf (CTD, KEGG, PubChem, DRUGBANK) """
def __init__(self, context): #triplestore):
super(ChemBioKS, self).__init__("chembio", context)
self.triplestore = TripleStore(self.url)
def query_chembio(self, query):
""" Execute and return the result of a SPARQL query. """
return self.triplestore.execute_query(query)
def get_exposure_conditions(self, chemicals):
""" Identify conditions (MeSH IDs) triggered by the specified stressor
agent ids (also MeSH IDs).
:param chemicals: List of IDs for substances of interest.
:type chemicals: list of MeSH IDs, eg. D052638
"""
id_list = ' '.join(
list(map(lambda d: "( mesh:{0} )".format(d), chemicals)))
text = self.triplestore.get_template ("ctd_gene_expo_disease").\
safe_substitute (chemicals=id_list)
results = self.triplestore.execute_query(text)
return list(
map(
lambda b: {
"chemical": b['chemical'].value,
"gene": b['gene'].value,
"pathway": b['kegg_pathway'].value,
"pathName": b['pathway_name'].value,
"pathID": b['pathway_id'].value,
"human": '(human)' in b['pathway_name'].value
}, results.bindings))
def get_drugs_by_condition(self, conditions):
""" Get drugs associated with a set of conditions.
:param conditions: Conditions to find associated drugs for.
:type conditions: List of MeSH IDs for conditions, eg.: D001249
"""
if not isinstance(conditions, list):
conditions = [conditions]
conditions = list(
map(lambda v: v.replace("MESH:", "mesh:"), conditions))
prefix = "mesh:"
if any(map(lambda v: v.startswith(prefix), conditions)):
prefix = ""
condition_list = ', '.join(
list(map(lambda d: " {0}{1} ".format(prefix, d), conditions)))
result = self.triplestore.query_template(
inputs={"diseaseIds": condition_list.lower()},
outputs=[
'drugID', 'drugGenericName', 'pubChemCID', 'diseasePMIDs'
],
template_text="""
prefix mesh: <http://bio2rdf.org/mesh:>
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
prefix db_resource: <http://chem2bio2rdf.org/drugbank/resource/>
select ?drugID ?drugGenericName ?diseasePMIDs ?ctdChemDis ?pubChemCID where {
values ( ?diseaseId ) { ( $diseaseIds ) }
?ctdChemDis ctd:cid ?pubChemCID;
ctd:diseaseid ?diseaseId;
ctd:pubmedids ?diseasePMIDs.
?dbInter db_resource:Name ?name ;
db_resource:DBID ?drugID .
?drugID db_resource:CID ?pubChemCID ;
db_resource:Generic_Name ?drugGenericName .
}""")
return result
def get_drugs_by_condition_graph(self, conditions):
drugs = self.get_drugs_by_condition(conditions.identifier)
results = []
for r in drugs:
edge = KEdge('c2b2r', 'conditionToDrug', {
'cid': r['pubChemCID'],
'pmids': r['diseasePMIDs']
})
node = KNode(
r['drugID'].split('/')[-1:][0],
#"http://chem2bio2rdf.org/drugbank/resource/drugbank_drug",
node_types.DRUG,
r['drugGenericName'])
results.append((edge, node))
#logger.debug ("chembio drugs by condition: {}".format (results))
return results
def get_genes_pathways_by_disease(self, diseases):
""" Get genes and pathways associated with specified conditions.
:param diseases: List of conditions designated by MeSH ID.
:return: Returns a list of dicts containing gene and path information.
"""
diseaseMeshIDList = ' '.join(
list(map(lambda d: "( mesh:{0} )".format(d), diseases)))
text = self.triplestore.get_template(
"genes_pathways_by_disease").safe_substitute(
diseaseMeshIDList=diseaseMeshIDList)
results = self.triplestore.execute_query(text)
return list(
map(
lambda b: {
"uniprotGene": b['uniprotGeneID'].value,
"keggPath": b['keggPath'].value,
"pathName": b['pathwayName'].value,
"human": '(human)' in b['pathwayName'].value
}, results.bindings))
def get_drug_gene_disease(self, disease_name, drug_name):
""" Identify targets and diseases assocaited with a drug name.
:param disease_name: MeSH name of a disease condition.
:type str: String
:param drug_name: Name of a drug.
:type str: String
"""
text = self.triplestore.get_template(
"drug_gene_disease").safe_substitute(diseaseName=disease_name,
drugName=drug_name)
results = self.triplestore.execute_query(text)
return list(
map(
lambda b: {
"uniprotSymbol": b['uniprotSym'].value,
"diseaseId": b['diseaseID'].value
}, results.bindings))
def pubchem_to_ncbigene(self, pubchemID):
result = self.triplestore.query_template(
inputs={"pubchemID": "pubchem:{}".format(pubchemID)},
outputs=[
'NCBIGene', 'meshID', 'interaction', 'interactionTypes',
'pubmedids'
],
template_text="""
prefix pubchem: <http://chem2bio2rdf.org/pubchem/resource/pubchem_compound/>
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
select distinct ?NCBIGene ?meshID ?interaction ?interactionTypes ?pubmedids where {
?ctdChemGene ctd:cid $pubchemID;
ctd:chemicalid ?meshID ;
ctd:geneid ?NCBIGene;
ctd:interaction ?interaction;
ctd:interactiontypes ?interactionTypes;
ctd:pubmedids ?pubmedids.
}""")
return list(
map(
lambda r: {
'NCBIGene': r['NCBIGene'],
'meshID': r['meshID'],
'interaction': r['interaction'],
'interactionTypes': r['interactionTypes'],
'pubmedids': r['pubmedids']
}, result))
def drug_name_to_gene_symbol(self, drug_name):
result = self.triplestore.query_template(
inputs={"drugName": drug_name},
outputs=['uniprotSym', 'pmids', 'drugID'],
template_text="""
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
prefix db_resource: <http://chem2bio2rdf.org/drugbank/resource/>
select ?drugGenericName ?pmids ?drugID ?uniprotSym where {
values ( ?drugName ) { ( "$drugName" ) }
?ctdChemGene ctd:cid ?pubChemCID;
ctd:pubmedids ?pmids;
ctd:gene ?uniprotSym .
?drugID db_resource:CID ?pubChemCID ;
db_resource:Generic_Name ?drugGenericName .
filter regex(lcase(str(?drugGenericName)), lcase(?drugName))
}""")
return list(
map(
lambda r: {
'uniprotSym': r['uniprotSym'],
'pmids': r.get('pmids', None),
'drugID': r['drugID']
}, result))
def drugname_to_pubchem(self, drug_name):
result = self.triplestore.query_template(
inputs={"drugName": drug_name},
outputs=['pubChemID', 'drugGenericName'],
template_text="""
prefix db_resource: <http://chem2bio2rdf.org/drugbank/resource/>
select distinct ?pubChemID ?drugGenericName where {
values ( ?drugName ) { ( "$drugName" ) }
?drugID db_resource:CID ?pubChemID ;
db_resource:Generic_Name ?drugGenericName .
filter regex(lcase(str(?drugGenericName)), lcase(?drugName))
}""")
return list(
map(
lambda r: {
'drugID': r['pubChemID'],
'drugName': r['drugGenericName']
}, result))
def gene_symbol_to_pathway(self, uniprot_symbol):
return self.triplestore.query_template(
inputs={"uniprotSymbol": uniprot_symbol},
outputs=["keggPath"],
template_text="""
prefix kegg: <http://chem2bio2rdf.org/kegg/resource/>
prefix pharmgkb: <http://chem2bio2rdf.org/pharmgkb/resource/>
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
select ?ctdGene ?uniprotID ?pathwayName ?keggPath where {
values ( ?ctdGene ) { ( <$uniprotSymbol> ) }
?keggPath kegg:protein ?uniprotID ; kegg:Pathway_name ?pathwayName .
?pharmGene pharmgkb:Symbol ?ctdGene ; pharmgkb:UniProt_Id ?uniprotID.
?ctdChemGene ctd:gene ?ctdGene.
} LIMIT 500
""")
def uniprot_to_hgnc(self, uniprot_symbol):
return self.triplestore.query_template(
inputs={"uniprotID": Text.un_curie(uniprot_symbol.identifier)},
outputs=["hgncID"],
template_text="""
prefix uniprot: <http://chem2bio2rdf.org/uniprot/resource/gene/>
prefix owl: <http://www.w3.org/2002/07/owl#>
prefix hgnc: <http://chem2bio2rdf.org/rdf/resource/hgnc/>
select distinct ?hgncID where {
values ( ?uniprotID ) { ( uniprot:${uniprotID} ) }
?uniprotID <http://www.w3.org/2002/07/owl#sameAs> ?hgncID.
filter ( strstarts (str(?hgncID), "http://bio2rdf.org/gene:"))
}
""")
def graph_uniprot_to_hgnc(self, uniprot_symbol):
result = self.uniprot_to_hgnc(uniprot_symbol)
return [(self.get_edge(r, predicate='synonym'),
KNode('HGNC:{0}'.format(r['hgncID'].split(':')[-1]),
node_types.GENE)) for r in result]
def graph_get_genes_by_disease(self, disease): #reasoner
disease = disease.identifier.split(':')[1].lower()
response = self.get_genes_pathways_by_disease([disease])
results = []
for r in response:
edge = KEdge('c2b2r', 'diseaseToGene', {'keggPath': r['keggPath']})
node = KNode(
"UNIPROT:{0}".format(r['uniprotGene'].split('/')[-1:][0]),
node_types.GENE)
results.append((edge, node))
return results
@cachier(stale_after=datetime.timedelta(days=20))
def graph_get_pathways_by_gene(self, gene): #reasoner
response = self.triplestore.query_template(
inputs={"gene": gene.identifier.split(':')[1].upper()},
outputs=['keggPath'],
template_text="""
prefix kegg: <http://chem2bio2rdf.org/kegg/resource/>
prefix drugbank: <http://chem2bio2rdf.org/drugbank/resource/>
prefix uniprot: <http://chem2bio2rdf.org/uniprot/resource/gene/>
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
prefix mesh: <http://bio2rdf.org/mesh:>
select ?drugGenericName ?uniprotGeneID ?pathwayName ?keggPath where {
?keggPath kegg:protein ?swissProtID ;
kegg:Pathway_name ?pathwayName .
?keggInter kegg:cid ?pubchemCID .
?dbInter drugbank:GeneBank_ID ?geneBankID ;
drugbank:SwissProt_ID ?swissProtID ;
drugbank:gene ?uniprotGeneID .
?drugID drugbank:CID ?pubchemCID ;
drugbank:Generic_Name ?drugGenericName .
?ctd_disease ctd:diseaseid ?diseaseID ;
ctd:cid ?pubchemCID .
values ( ?uniprotGeneID ) {
( uniprot:$gene )
}
} LIMIT 2000""")
results = []
for r in response:
edge = KEdge('c2b2r', 'geneToPathway', {})
node = KNode("KEGG:{0}".format(r['keggPath'].split('/')[-1:][0]),
node_types.PATHWAY)
results.append((edge, node))
return results
def graph_drugname_to_gene_symbol(self, drug_name_node):
drug_name = Text.un_curie(drug_name_node.identifier)
response = self.drug_name_to_gene_symbol(drug_name)
results = []
for r in response:
edge = self.get_edge(r, predicate="targets")
node = KNode("UNIPROT:{0}".format(Text.path_last(r['uniprotSym'])),
node_types.GENE)
results.append((edge, node))
return results
def graph_name_to_drugbank(self, drug_name_node):
drug_name = Text.un_curie(drug_name_node.identifier)
response = self.drug_name_to_gene_symbol(drug_name)
results = []
for r in response:
edge = self.get_edge(r, predicate="drugname")
node = KNode ("DRUGBANK:{0}".format (Text.path_last (r['drugID'])), \
node_types.DRUG, \
label=r['drugName'])
results.append((edge, node))
return results
def graph_get_pathways_by_gene(self, gene): #reasoner
response = self.triplestore.query_template(
inputs={"gene": gene.identifier.split(':')[1].upper()},
outputs=['keggPath'],
template_text="""
prefix kegg: <http://chem2bio2rdf.org/kegg/resource/>
prefix drugbank: <http://chem2bio2rdf.org/drugbank/resource/>
prefix uniprot: <http://chem2bio2rdf.org/uniprot/resource/gene/>
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
prefix mesh: <http://bio2rdf.org/mesh:>
select ?drugGenericName ?uniprotGeneID ?pathwayName ?keggPath where {
?keggPath kegg:protein ?swissProtID ;
kegg:Pathway_name ?pathwayName .
?keggInter kegg:cid ?pubchemCID .
?dbInter drugbank:GeneBank_ID ?geneBankID ;
drugbank:SwissProt_ID ?swissProtID ;
drugbank:gene ?uniprotGeneID .
?drugID drugbank:CID ?pubchemCID ;
drugbank:Generic_Name ?drugGenericName .
?ctd_disease ctd:diseaseid ?diseaseID ;
ctd:cid ?pubchemCID .
values ( ?uniprotGeneID ) {
( uniprot:$gene )
}
} LIMIT 2000""")
results = []
for r in response:
edge = KEdge('c2b2r', 'geneToPathway', {})
node = KNode("KEGG:{0}".format(r['keggPath'].split('/')[-1:][0]),
node_types.PATHWAY)
results.append((edge, node))
return results
def graph_drugbank_to_uniprot(self, drugbank):
response = self.triplestore.query_template(inputs={
"drugID":
"DB{0}".format(Text.un_curie(drugbank.identifier))
},
outputs=["uniprotGeneID"],
template_text="""
prefix drugbank: <http://chem2bio2rdf.org/drugbank/resource/>
prefix drugbank_drug: <http://chem2bio2rdf.org/drugbank/resource/drugbank_drug/>
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
select distinct ?uniprotGeneID where {
values ( ?drugID ) { ( drugbank_drug:${drugID} ) }
?dbInter drugbank:GeneBank_ID ?geneBankID ;
drugbank:gene ?uniprotGeneID .
?drugID drugbank:CID ?pubchemCID ;
drugbank:Generic_Name ?drugGenericName .
?ctd_disease ctd:diseaseid ?diseaseID ;
ctd:cid ?pubchemCID .
}""")
return [
(self.get_edge(r, predicate='targets'),
KNode("UNIPROT:{0}".format(r['uniprotGeneID'].split('/')[-1:][0]),
node_types.GENE)) for r in response
]
def graph_diseasename_to_uniprot(self, disease):
results = []
response = self.triplestore.query_template(
inputs={"diseaseName": Text.un_curie(disease.identifier)},
outputs=["pubChemCID"],
template_text="""
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
select distinct ?pubChemCID where {
values ( ?diseaseName ) { ( "$diseaseName" ) }
?ctdChemDis ctd:cid ?pubChemCID;
ctd:diseasename ?diseaseNameRec.
filter regex(lcase(str(?diseaseNameRec)), lcase(?diseaseName))
} LIMIT 1""")
if len(response) > 0: # This is a disease.
response = self.triplestore.query_template(
inputs={"diseaseName": Text.un_curie(disease.identifier)},
outputs=["disPmids", "chemPmids", "uniprotSym"],
template_text="""
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
select ?disPmids ?ctdChemDis ?chemPmids ?uniprotSym ?diseaseId where {
values ( ?diseaseName ) { ( "$diseaseName" ) }
?ctdChemGene ctd:cid ?pubChemCID;
ctd:pubmedids ?chemPmids;
ctd:gene ?uniprotSym.
?ctdChemDis ctd:cid ?pubChemCID;
ctd:diseaseid ?diseaseId;
ctd:diseasename ?diseaseNameRec;
ctd:pubmedids ?disPmids.
filter regex(lcase(str(?diseaseNameRec)), lcase(?diseaseName))
} LIMIT 500""")
for r in response:
chemPmids = r['chemPmids']
disPmids = r['disPmids']
pmids = chemPmids + "|" + disPmids
edge = self.get_edge(r, predicate='caused_by', pmids=pmids),
node = KNode(
"UNIPROT:{0}".format(r['uniprotSym'].split('/')[-1:][0]),
node_types.GENE)
results.append((edge, node))
return results
def graph_diseaseid_to_uniprot(self, drugbank):
print(drugbank.identifier.lower())
response = self.triplestore.query_template(
inputs={"diseaseID": drugbank.identifier.lower()},
outputs=["uniprotGeneID"],
template_text="""
prefix drugbank: <http://chem2bio2rdf.org/drugbank/resource/>
prefix drugbank_drug: <http://chem2bio2rdf.org/drugbank/resource/drugbank_drug/>
prefix ctd: <http://chem2bio2rdf.org/ctd/resource/>
prefix mesh.disease: <http://bio2rdf.org/mesh:>
select distinct ?uniprotGeneID where {
values ( ?diseaseID ) { ( $diseaseID ) }
?dbInter drugbank:gene ?uniprotGeneID .
?drugID drugbank:CID ?pubchemCID.
?ctd_disease ctd:diseaseid ?diseaseID ;
ctd:cid ?pubchemCID .
}""")
return [
(self.get_edge(r, predicate='targets'),
KNode("UNIPROT:{0}".format(r['uniprotGeneID'].split('/')[-1:][0]),
node_types.GENE)) for r in response
]
def graph_drugname_to_pubchem(self, drugname_node):
drug_name = Text.un_curie(drugname_node.identifier)
response = self.drugname_to_pubchem(drug_name)
return [ (self.get_edge( r, predicate='drugname_to_pubchem'), \
KNode( "PUBCHEM:{}".format( r['drugID'].split('/')[-1]), node_types.DRUG, label=r['drugName'])) for r in response ]
# 'NCBIGene' : r['NCBIGene'],
# 'meshID' : r['meshID'],
# 'interaction': r['interaction'],
# 'interactionTypes': r['interactionTypes']
# 'pubmedids' : r['pubmedids']
def graph_pubchem_to_ncbigene(self, pubchem_node):
#The compound mesh coming back from here is very out of date. Ignore.
pubchemid = Text.un_curie(pubchem_node.identifier)
response = self.pubchem_to_ncbigene(pubchemid)
retvals = []
for r in response:
props = {}
props['interaction'] = r['interaction']
props['interactionTypes'] = r['interactionTypes']
props['publications'] = r['pubmedids'].split('|')
retvals.append((self.get_edge(props,
predicate='pubchem_to_ncbigene'),
KNode("NCBIGene:{}".format(r['NCBIGene']),
node_types.GENE)))
return retvals
def __init__(self, uri="http://id.nlm.nih.gov/mesh/sparql"):
self.triplestore = TripleStore (uri)
class OntologicalHeirarchy(Service):
"""
Service that makes call to uberongraph to resolve subclass relationships between ontological terms
"""
def __init__(self):
self.url = "https://stars-app.renci.org/uberongraph/sparql"
self.triplestore = TripleStore(self.url)
self.prefix_set = {
node_types.DISEASE_OR_PHENOTYPIC_FEATURE: ['HP', 'MONDO'],
node_types.CELLULAR_COMPONENT: ['CL'],
node_types.BIOLOGICAL_PROCESS_OR_ACTIVITY: ['GO'],
node_types.ANATOMICAL_ENTITY: ['UBERON'],
node_types.CHEMICAL_SUBSTANCE: ['CHEBI']
}
self.root_uris = {
node_types.ANATOMICAL_ENTITY:
"<http://purl.obolibrary.org/obo/UBERON_0001062>",
node_types.DISEASE:
"<http://purl.obolibrary.org/obo/MONDO_0000001>",
node_types.MOLECULAR_ACTIVITY:
"<http://purl.obolibrary.org/obo/GO_0003674>",
node_types.BIOLOGICAL_PROCESS:
"<http://purl.obolibrary.org/obo/GO_0008150>",
node_types.CHEMICAL_SUBSTANCE:
"<http://purl.obolibrary.org/obo/CHEBI_24431>",
node_types.PHENOTYPIC_FEATURE:
"<http://purl.obolibrary.org/obo/HP_0000118>",
node_types.CELL:
"<http://purl.obolibrary.org/obo/CL_0000000>",
node_types.CELLULAR_COMPONENT:
"<http://purl.orolibrary.org/obo/GO_0005575>"
}
obo_prefixes = '\n'.join([
f'PREFIX {pref}: <http://purl.obolibrary.org/obo/{pref}_>'
for pref in set(
reduce(lambda x, y: x + y, self.prefix_set.values(), []))
])
self.query = f"""
{obo_prefixes}
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
select ?parent_id ?parent_label ?child_id ?child_label
where {{
?parent_id rdfs:subClassOf $root_uri .
?child_id rdfs:subClassOf ?parent_id.
OPTIONAL {{
?parent_id rdfs:label ?parent_label.
?child_id rdfs:label ?child_label.
}}
}}
"""
rosetta = Rosetta()
self.wdg = WriterDelegator(rosetta)
def runner(self):
for node_type, root_iri in self.root_uris.items():
nodes, edges = self.term_get_ancestors(node_type, root_iri)
for index, n in enumerate(nodes):
self.wdg.write_node(n, annotate=False)
if ((index / len(nodes)) * 100) % 10 == 0:
print((index / len(nodes)) * 100, '% complete')
for index, e in enumerate(edges):
self.wdg.write_edge(e)
if index % 100 == 0:
self.wdg.flush()
if ((index / len(edges)) * 100) % 10 == 0:
print((index / len(edges)) * 100, '% complete')
return
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
class OntologicalHeirarchy(Service):
"""
Service that makes call to uberongraph to resolve subclass relationships between ontological terms
"""
def __init__(self, context):
super(OntologicalHeirarchy, self).__init__("ontological_hierarchy",
context)
self.triplestore = TripleStore(self.url)
self.prefix_set = {
node_types.DISEASE_OR_PHENOTYPIC_FEATURE: ['HP', 'MONDO'],
node_types.CELLULAR_COMPONENT: ['CL'],
node_types.BIOLOGICAL_PROCESS_OR_ACTIVITY: ['GO'],
node_types.ANATOMICAL_ENTITY: ['UBERON'],
node_types.CHEMICAL_SUBSTANCE: ['CHEBI']
}
self.root_uris = {
node_types.ANATOMICAL_ENTITY:
"<http://purl.obolibrary.org/obo/UBERON_0001062>",
node_types.DISEASE:
"<http://purl.obolibrary.org/obo/MONDO_0000001>",
node_types.MOLECULAR_ACTIVITY:
"<http://purl.obolibrary.org/obo/GO_0003674>",
node_types.BIOLOGICAL_PROCESS:
"<http://purl.obolibrary.org/obo/GO_0008150>",
node_types.CHEMICAL_SUBSTANCE:
"<http://purl.obolibrary.org/obo/CHEBI_24431>",
node_types.PHENOTYPIC_FEATURE:
"<http://purl.obolibrary.org/obo/HP_0000118>",
node_types.CELL:
"http://purl.obolibrary.org/obo/CL_0000000",
node_types.CELLULAR_COMPONENT:
"http://purl.orolibrary.org/obo/GO_0005575"
}
obo_prefixes = '\n'.join([
f'PREFIX {pref}: <http://purl.obolibrary.org/obo/{pref}_>'
for pref in set(
reduce(lambda x, y: x + y, self.prefix_set.values(), []))
])
self.query = f"""
{obo_prefixes}
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
select distinct ?parent_id ?label
where {{
graph <http://reasoner.renci.org/ontology/closure> {{
$child_curie rdfs:subClassOf ?parent_id .
?parent_id rdfs:subClassOf $root_uri .
}}
graph <http://reasoner.renci.org/ontology>{{
?parent_id rdfs:label ?label.
}}
}}
"""
def term_get_ancestors(self, child_node):
root_uri = self.root_uris.get(child_node.type, None)
if not root_uri:
return []
###
# Query does have an upper bound so for ontologies that start from
#
# Step 1 get prefixes that are supported for input node
curie_set = set()
for node_type in child_node.export_labels:
ps = self.prefix_set.get(node_type, [])
for prefix in ps:
synonyms = child_node.get_synonyms_by_prefix(prefix)
curie_set.update(synonyms)
# Step 2 get parents for those curies we support from uberon graph
outputs = []
for curie in curie_set:
results = self.triplestore.query_template(
template_text=self.query,
inputs={
'child_curie': curie,
'root_uri': root_uri
},
outputs=['parent_id', 'label'])
for row in results:
# Output type would be same as input type?
ancestor_node = KNode(Text.obo_to_curie(row['parent_id']),
name=row['label'],
type=child_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)
outputs.append((edge, ancestor_node))
return outputs
def __init__(self, context, url):
super(MeshKS, self).__init__("mesh", context)
self.triplestore = TripleStore(self.url)
class OmniCorp(Service):
def __init__(self, context): #triplestore):
super(OmniCorp, self).__init__("omnicorp", context)
self.triplestore = TripleStore(self.url)
self.prefix_to_uri = {
'UBERON': 'http://purl.obolibrary.org/obo/UBERON_',
'BSPO': 'http://purl.obolibrary.org/obo/BSPO_',
'PATO': 'http://purl.obolibrary.org/obo/PATO_',
'GO': 'http://purl.obolibrary.org/obo/GO_',
'MONDO': 'http://purl.obolibrary.org/obo/MONDO_',
'HP': 'http://purl.obolibrary.org/obo/HP_',
'ENVO:': 'http://purl.obolibrary.org/obo/ENVO_',
'OBI': 'http://purl.obolibrary.org/obo/OBI_',
'CL': 'http://purl.obolibrary.org/obo/CL_',
'SO': 'http://purl.obolibrary.org/obo/SO_',
'CHEBI': 'http://purl.obolibrary.org/obo/CHEBI_',
'HGNC': 'http://identifiers.org/hgnc/HGNC:',
'MESH': 'http://id.nlm.nih.gov/mesh/'
}
def get_omni_identifier(self, node):
#Let's start with just the 'best' identifier
identifier = node.id
prefix = Text.get_curie(node.id)
if prefix not in self.prefix_to_uri:
logger.warn("What kinda tomfoolery is this?")
logger.warn(f"{node.id} {node.type}")
logger.warn(f"{node.synonyms}")
return None
oident = f'{self.prefix_to_uri[prefix]}{Text.un_curie(node.id)}'
return oident
def query_omnicorp(self, query):
""" Execute and return the result of a SPARQL query. """
return self.triplestore.execute_query(query)
def sparql_get_all_shared_pmids(self, identifier_list):
text = """
PREFIX dct: <http://purl.org/dc/terms/>
SELECT DISTINCT ?pubmed ?term1 ?term2
WHERE {
hint:Query hint:analytic true .
VALUES ?term1 $id_list_a
VALUES ?term2 $id_list_b
?pubmed dct:references ?term1 .
?pubmed dct:references ?term2 .
FILTER(STR(?term1) < STR(?term2))
}
"""
start = datetime.datetime.now()
results = self.triplestore.query_template(
inputs={
'id_list_a': identifier_list,
'id_list_b': identifier_list
},
outputs=['term1', 'term2', 'pubmed'],
template_text=text,
post=True)
end = datetime.datetime.now()
logger.debug(f'Completed in: {end-start}')
return results
def sparql_count_pmids(self, identifier):
text = """
PREFIX dct: <http://purl.org/dc/terms/>
SELECT (COUNT(DISTINCT ?pubmed) as ?count)
WHERE {
hint:Query hint:analytic true .
?pubmed dct:references <$identifier> .
}
"""
results = self.triplestore.query_template(
inputs={'identifier': identifier},
outputs=['count'],
template_text=text,
)
return results
def sparql_get_shared_pmids(self, identifier_a, identifier_b):
text = """
PREFIX dct: <http://purl.org/dc/terms/>
SELECT DISTINCT ?pubmed
WHERE {
hint:Query hint:analytic true .
?pubmed dct:references <$id_a> .
?pubmed dct:references <$id_b> .
}
"""
results = self.triplestore.query_template(inputs={
'id_a': identifier_a,
'id_b': identifier_b
},
outputs=['pubmed'],
template_text=text,
post=True)
return results
def get_all_shared_pmids(self, nodes):
oiddict = {self.get_omni_identifier(n): n for n in nodes}
oids = [
f'<{x}>' for x in filter(lambda n: n is not None, oiddict.keys())
]
oidsstring = '{ ' + ' '.join(oids) + '}'
results = self.sparql_get_all_shared_pmids(oidsstring)
pubmeds = defaultdict(list)
for r in results:
k = (oiddict[r['term1']], oiddict[r['term2']])
pubmeds[k].append(f"PMID:{r['pubmed'].split('/')[-1]}")
for i, node_i in enumerate(nodes):
for node_j in nodes[:i]:
k_ij = (node_i, node_j)
k_ji = (node_j, node_i)
if k_ij not in pubmeds and k_ji not in pubmeds:
pubmeds[k_ij] = []
return pubmeds
def call_with_retries(self, fnc, args):
done = False
ntries = 0
maxtries = 100
rest_time = 10 #seconds
start = datetime.datetime.now()
while not done and ntries < maxtries:
try:
result = fnc(*args)
done = True
except:
logger.warn("OmniCorp error, retrying")
time.sleep(rest_time)
ntries += 1
if not done:
return None
else:
end = datetime.datetime.now()
logger.debug(f'Total call ntries: {ntries}, time: {end-start}')
return result
def count_pmids(self, node):
identifier = self.get_omni_identifier(node)
if identifier is None:
return 0
res = self.call_with_retries(self.sparql_count_pmids, [identifier])
if res is None:
return None
else:
logger.debug(f"Returned {res[0]['count']}")
return res[0]['count']
def get_shared_pmids(self, node1, node2):
id1 = self.get_omni_identifier(node1)
id2 = self.get_omni_identifier(node2)
if id1 is None or id2 is None:
return []
done = False
ntries = 0
pmids = self.call_with_retries(self.sparql_get_shared_pmids,
[id1, id2])
if pmids is None:
logger.error("OmniCorp gave up")
return None
return [p['pubmed'] for p in pmids]
class UberonGraphKS(Service):
"""A knowledge source created by 1) Combining cell ontology, uberon, and
HPO, 2) Reasoning over the total graph to realize many implicit edges.
Created by Jim Balhoff"""
def __init__(self, context): #triplestore):
super(UberonGraphKS, self).__init__("uberongraph", context)
self.triplestore = TripleStore(self.url)
def query_uberongraph(self, query):
""" Execute and return the result of a SPARQL query. """
return self.triplestore.execute_query(query)
def cell_get_cellname(self, cell_identifier):
""" Identify label for a cell type
:param cell: CL identifier for cell type
"""
text = """
prefix CL: <http://purl.obolibrary.org/obo/CL_>
select distinct ?cellLabel
from <http://reasoner.renci.org/nonredundant>
from <http://example.org/uberon-hp-cl.ttl>
where {
$cellID rdfs:label ?cellLabel .
}
"""
results = self.triplestore.query_template(
inputs = { 'cellID': cell_identifier }, \
outputs = [ 'cellLabel' ], \
template_text = text \
)
return results
def get_anatomy_parts(self, anatomy_identifier):
"""Given an UBERON id, find other UBERONS that are parts of the query"""
if anatomy_identifier.startswith('http'):
anatomy_identifier = Text.obo_to_curie(anatomy_identifier)
text = """
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
prefix BFO: <http://purl.obolibrary.org/obo/BFO_>
select distinct ?part ?partlabel
from <http://reasoner.renci.org/nonredundant>
from <http://example.org/uberon-hp-cl.ttl>
where {
$anatomy_id BFO:0000051 ?part .
graph <http://reasoner.renci.org/redundant> {
?part rdfs:subClassOf UBERON:0001062 .
}
?part rdfs:label ?partlabel .
}
"""
results = self.triplestore.query_template(
inputs = { 'anatomy_id': anatomy_identifier }, \
outputs = [ 'part', 'partlabel' ], \
template_text = text \
)
for result in results:
result['curie'] = Text.obo_to_curie(result['part'])
return results
def anatomy_to_cell(self, anatomy_identifier):
""" Identify anatomy terms related to cells.
:param cell: CL identifier for cell type
"""
text = """
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
prefix CL: <http://purl.obolibrary.org/obo/CL_>
prefix BFO: <http://purl.obolibrary.org/obo/BFO_>
select distinct ?cellID ?cellLabel
from <http://reasoner.renci.org/nonredundant>
from <http://example.org/uberon-hp-cl.ttl>
where {
graph <http://reasoner.renci.org/redundant> {
?cellID rdfs:subClassOf CL:0000000 .
?cellID BFO:0000050 $anatomyID .
}
?cellID rdfs:label ?cellLabel .
}
"""
results = self.triplestore.query_template(
inputs = { 'anatomyID': anatomy_identifier }, \
outputs = [ 'cellID', 'cellLabel' ], \
template_text = text \
)
return results
def cell_to_anatomy(self, cell_identifier):
""" Identify anatomy terms related to cells.
:param cell: CL identifier for cell type
"""
text = """
prefix CL: <http://purl.obolibrary.org/obo/CL_>
prefix BFO: <http://purl.obolibrary.org/obo/BFO_>
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
select distinct ?anatomyID ?anatomyLabel
from <http://reasoner.renci.org/nonredundant>
from <http://example.org/uberon-hp-cl.ttl>
where {
graph <http://reasoner.renci.org/redundant> {
?anatomyID rdfs:subClassOf UBERON:0001062 .
$cellID BFO:0000050 ?anatomyID .
}
?anatomyID rdfs:label ?anatomyLabel .
}
"""
results = self.triplestore.query_template(
inputs = { 'cellID': cell_identifier }, \
outputs = [ 'anatomyID', 'anatomyLabel' ], \
template_text = text \
)
return results
def phenotype_to_anatomy(self, hp_identifier):
""" Identify anatomy terms related to cells.
:param cell: HP identifier for phenotype
"""
#The subclassof uberon:0001062 ensures that the result
#is an anatomical entity.
text = """
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
prefix HP: <http://purl.obolibrary.org/obo/HP_>
prefix part_of: <http://purl.obolibrary.org/obo/BFO_0000050>
prefix has_part: <http://purl.obolibrary.org/obo/BFO_0000051>
prefix depends_on: <http://purl.obolibrary.org/obo/RO_0002502>
prefix phenotype_of: <http://purl.obolibrary.org/obo/UPHENO_0000001>
select distinct ?anatomy_id ?anatomy_label ?input_label
from <http://reasoner.renci.org/nonredundant>
from <http://example.org/uberon-hp-cl.ttl>
where {
graph <http://reasoner.renci.org/redundant> {
?anatomy_id rdfs:subClassOf UBERON:0001062 .
}
?anatomy_id rdfs:label ?anatomy_label .
graph <http://reasoner.renci.org/nonredundant> {
?phenotype phenotype_of: ?anatomy_id .
}
graph <http://reasoner.renci.org/redundant> {
$HPID rdfs:subClassOf ?phenotype .
}
$HPID rdfs:label ?input_label .
}
"""
results = self.triplestore.query_template(
inputs = { 'HPID': hp_identifier }, \
outputs = [ 'anatomy_id', 'anatomy_label', 'input_label'],\
template_text = text \
)
return results
def anatomy_to_phenotype(self, uberon_id):
text = """
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
prefix HP: <http://purl.obolibrary.org/obo/HP_>
prefix part_of: <http://purl.obolibrary.org/obo/BFO_0000050>
prefix has_part: <http://purl.obolibrary.org/obo/BFO_0000051>
prefix depends_on: <http://purl.obolibrary.org/obo/RO_0002502>
prefix phenotype_of: <http://purl.obolibrary.org/obo/UPHENO_0000001>
select distinct ?pheno_id ?anatomy_label ?pheno_label
from <http://reasoner.renci.org/nonredundant>
from <http://example.org/uberon-hp-cl.ttl>
where {
$UBERONID rdfs:label ?anatomy_label .
graph <http://reasoner.renci.org/nonredundant> {
?phenotype phenotype_of: $UBERONID .
}
graph <http://reasoner.renci.org/redundant> {
?pheno_id rdfs:subClassOf ?phenotype .
}
?pheno_id rdfs:label ?pheno_label .
}
"""
#The subclassof uberon:0001062 ensures that the result
#is an anatomical entity.
results = self.triplestore.query_template(
inputs = { 'UBERONID': uberon_id }, \
outputs = [ 'pheno_id', 'anatomy_label', 'pheno_label'],\
template_text = text \
)
return results
def get_anatomy_by_cell_graph(self, cell_node):
anatomies = self.cell_to_anatomy(cell_node.id)
results = []
predicate = LabeledID(identifier='BFO:0000050', label='part_of')
for r in anatomies:
anatomy_node = KNode(Text.obo_to_curie(r['anatomyID']),
type=node_types.ANATOMY,
name=r['anatomyLabel'])
edge = self.create_edge(cell_node, anatomy_node,
'uberongraph.get_anatomy_by_cell_graph',
cell_node.id, predicate)
results.append((edge, anatomy_node))
return results
def get_cell_by_anatomy_graph(self, anatomy_node):
cells = self.anatomy_to_cell(anatomy_node.id)
results = []
predicate = LabeledID(identifier='BFO:0000050', label='part_of')
for r in cells:
cell_node = KNode(Text.obo_to_curie(r['cellID']),
type=node_types.CELL,
name=r['cellLabel'])
edge = self.create_edge(cell_node, anatomy_node,
'uberongraph.get_cell_by_anatomy_graph',
anatomy_node.id, predicate)
results.append((edge, cell_node))
return results
def create_phenotype_anatomy_edge(self, node_id, node_label, input_id,
phenotype_node):
predicate = LabeledID(identifier='GAMMA:0000002',
label='inverse of has phenotype affecting')
anatomy_node = KNode(Text.obo_to_curie(node_id),
type=node_types.ANATOMY,
name=node_label)
edge = self.create_edge(anatomy_node, phenotype_node,
'uberongraph.get_anatomy_by_phenotype_graph',
input_id, predicate)
#node.name = node_label
return edge, anatomy_node
def create_anatomy_phenotype_edge(self, node_id, node_label, input_id,
anatomy_node):
predicate = LabeledID(identifier='GAMMA:0000002',
label='inverse of has phenotype affecting')
phenotype_node = KNode(Text.obo_to_curie(node_id),
type=node_types.PHENOTYPE,
name=node_label)
edge = self.create_edge(anatomy_node, phenotype_node,
'uberongraph.get_phenotype_by_anatomy_graph',
input_id, predicate)
#node.name = node_label
return edge, phenotype_node
def get_anatomy_by_phenotype_graph(self, phenotype_node):
results = []
for curie in phenotype_node.get_synonyms_by_prefix('HP'):
anatomies = self.phenotype_to_anatomy(curie)
for r in anatomies:
edge, node = self.create_phenotype_anatomy_edge(
r['anatomy_id'], r['anatomy_label'], curie, phenotype_node)
if phenotype_node.name is None:
phenotype_node.name = r['input_label']
results.append((edge, node))
#These tend to be very high level terms. Let's also get their parts to
#be more inclusive.
#TODO: there ought to be a more principled way to take care of this, but
#it highlights the uneasy relationship between the high level world of
#smartapi and the low-level sparql-vision.
part_results = self.get_anatomy_parts(r['anatomy_id'])
for pr in part_results:
pedge, pnode = self.create_phenotype_anatomy_edge(
pr['part'], pr['partlabel'], curie, phenotype_node)
results.append((pedge, pnode))
return results
def get_phenotype_by_anatomy_graph(self, anatomy_node):
results = []
for curie in anatomy_node.get_synonyms_by_prefix('UBERON'):
phenotypes = self.anatomy_to_phenotype(curie)
for r in phenotypes:
edge, node = self.create_anatomy_phenotype_edge(
r['pheno_id'], r['pheno_label'], curie, anatomy_node)
if anatomy_node.name is None:
anatomy_node.name = r['anatomy_label']
results.append((edge, node))
return results
class UberonGraphKS(Service):
"""A knowledge source created by 1) Combining cell ontology, uberon, and
HPO, 2) Reasoning over the total graph to realize many implicit edges.
Created by Jim Balhoff"""
def __init__(self, context): #triplestore):
super(UberonGraphKS, self).__init__("uberongraph", context)
self.triplestore = TripleStore(self.url)
#TODO: Pull this from the biolink model?
self.class_defs = {
node_types.CELL: 'CL:0000000',
node_types.ANATOMICAL_ENTITY: 'UBERON:0001062',
node_types.BIOLOGICAL_PROCESS: 'GO:0008150',
node_types.MOLECULAR_ACTIVITY: 'GO:0003674',
node_types.CHEMICAL_SUBSTANCE: 'CHEBI:24431',
node_types.DISEASE: 'MONDO:0000001',
node_types.PHENOTYPIC_FEATURE: 'UPHENO:0001002'
}
def query_uberongraph(self, query):
""" Execute and return the result of a SPARQL query. """
return self.triplestore.execute_query(query)
def get_edges(self, source_type, obj_type):
"""Given an UBERON id, find other UBERONS that are parts of the query"""
text = """
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
prefix CL: <http://purl.obolibrary.org/obo/CL_>
prefix GO: <http://purl.obolibrary.org/obo/GO_>
prefix CHEBI: <http://purl.obolibrary.org/obo/CHEBI_>
prefix MONDO: <http://purl.obolibrary.org/obo/MONDO_>
prefix UPHENO: <http://purl.obolibrary.org/obo/UPHENO_>
prefix BFO: <http://purl.obolibrary.org/obo/BFO_>
select distinct ?p ?pLabel
from <http://reasoner.renci.org/ontology>
where {
graph <http://reasoner.renci.org/redundant> {
?sourceID ?p ?objID .
}
graph <http://reasoner.renci.org/ontology/closure> {
?sourceID rdfs:subClassOf $sourcedefclass .
}
graph <http://reasoner.renci.org/ontology/closure> {
?objID rdfs:subClassOf $objdefclass .
hint:Prior hint:runFirst true .
}
?p rdfs:label ?pLabel .
}
"""
results = self.triplestore.query_template(
inputs = { 'sourcedefclass': self.class_defs[source_type], 'objdefclass': self.class_defs[obj_type] }, \
outputs = [ 'p', 'pLabel' ], \
template_text = text \
)
return results
def get_label(self, identifier):
obo_id = Text.curie_to_obo(identifier)
text = """
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
select distinct ?label
from <http://reasoner.renci.org/ontology>
where {
$obo_id rdfs:label ?label .
}
"""
results = self.triplestore.query_template(inputs={'obo_id': obo_id},
outputs=['label'],
template_text=text)
if len(results) < 1:
return ''
return results[0]['label']
def cell_get_cellname(self, cell_identifier):
""" Identify label for a cell type
:param cell: CL identifier for cell type
"""
text = """
prefix CL: <http://purl.obolibrary.org/obo/CL_>
select distinct ?cellLabel
from <http://reasoner.renci.org/nonredundant>
from <http://reasoner.renci.org/ontology>
where {
$cellID rdfs:label ?cellLabel .
}
"""
results = self.triplestore.query_template(
inputs = { 'cellID': cell_identifier }, \
outputs = [ 'cellLabel' ], \
template_text = text \
)
return results
def get_anatomy_parts(self, anatomy_identifier):
"""Given an UBERON id, find other UBERONS that are parts of the query"""
anatomy_identifier = f"<{anatomy_identifier}>"
text = """
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
prefix BFO: <http://purl.obolibrary.org/obo/BFO_>
select distinct ?part ?partlabel
from <http://reasoner.renci.org/nonredundant>
from <http://reasoner.renci.org/ontology>
where {
$anatomy_id BFO:0000051 ?part .
graph <http://reasoner.renci.org/ontology/closure> {
?part rdfs:subClassOf UBERON:0001062 .
}
?part rdfs:label ?partlabel .
}
"""
results = self.triplestore.query_template(
inputs = { 'anatomy_id': anatomy_identifier }, \
outputs = [ 'part', 'partlabel' ], \
template_text = text \
)
for result in results:
result['curie'] = Text.obo_to_curie(result['part'])
return results
def get_neighbor(self, input_id, output_type, subject=True):
parents = {
node_types.ANATOMICAL_ENTITY:
"<http://purl.obolibrary.org/obo/UBERON_0001062>",
node_types.DISEASE:
"<http://purl.obolibrary.org/obo/MONDO_0000001>",
node_types.MOLECULAR_ACTIVITY:
"<http://purl.obolibrary.org/obo/GO_0003674>",
node_types.BIOLOGICAL_PROCESS:
"<http://purl.obolibrary.org/obo/GO_0008150>",
node_types.CHEMICAL_SUBSTANCE:
"<http://purl.obolibrary.org/obo/CHEBI_24431>",
node_types.PHENOTYPIC_FEATURE:
"<http://purl.obolibrary.org/obo/HP_0000118>"
}
pref = Text.get_curie(input_id)
obo_prefix = f'PREFIX {pref}: <http://purl.obolibrary.org/obo/{pref}_>'
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
""" + obo_prefix + """
select distinct ?output_id ?output_label ?p ?pLabel
from <http://reasoner.renci.org/nonredundant>
from <http://reasoner.renci.org/ontology>
where {
graph <http://reasoner.renci.org/nonredundant> {
"""
if subject:
text += ' $input_id ?p ?output_id .'
else:
text += ' ?output_id ?p $input_id .'
text += """
}
graph <http://reasoner.renci.org/ontology/closure> {
?output_id rdfs:subClassOf $parent .
}
?output_id rdfs:label ?output_label .
?p rdfs:label ?pLabel .
}
"""
results = self.triplestore.query_template(
inputs={
'input_id': input_id,
'parent': parents[output_type]
},
outputs=['output_id', 'output_label', 'p', 'pLabel'],
template_text=text)
return results
def anatomy_to_anatomy(self, identifier):
results = {'subject': [], 'object': []}
for direction,query in \
(('subject',' ?input_id ?p ?output_id .'),
('object',' ?output_id ?p ?input_id .')):
text=""" PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
select distinct ?output_id ?output_label ?p ?pLabel
from <http://reasoner.renci.org/nonredundant>
from <http://reasoner.renci.org/ontology>
where {
graph <http://reasoner.renci.org/redundant> {
""" + query + \
"""
}
graph <http://reasoner.renci.org/ontology/closure> {
?output_id rdfs:subClassOf UBERON:0001062 .
}
?output_id rdfs:label ?output_label .
?p rdfs:label ?pLabel .
}
"""
results[direction] += self.triplestore.query_template(
inputs={'input_id': identifier},
outputs=['output_id', 'output_label', 'p', 'pLabel'],
template_text=text)
return results
def anatomy_to_go(self, anatomy_identifier):
""" Identify process and functions related to anatomical terms (anatomy, cell, components).
"""
#This is a bit messy, but we need to do 4 things. We are looking for go terms
# that are either biological processes or activities and we are looking for predicates
# that point either direction.
results = {'subject': [], 'object': []}
for goParent in ('GO:0008150', 'GO:0003674'):
for direction, query in (('subject', ' $anatID ?p ?goID'),
('object', ' ?goID ?p $anatID')):
text = """
prefix GO: <http://purl.obolibrary.org/obo/GO_>
select distinct ?goID ?goLabel ?p ?pLabel
from <http://reasoner.renci.org/nonredundant>
from <http://reasoner.renci.org/ontology>
where {
graph <http://reasoner.renci.org/redundant> {
""" + query + """
}
graph <http://reasoner.renci.org/ontology/closure> {
?goID rdfs:subClassOf $goParent .
}
?goID rdfs:label ?goLabel .
?p rdfs:label ?pLabel
}
"""
results[direction] += self.triplestore.query_template(
inputs = { 'anatID': anatomy_identifier, 'goParent': goParent }, \
outputs = [ 'goID', 'goLabel', 'p', 'pLabel' ], \
template_text = text \
)
return results
def go_to_anatomy(self, input_identifier):
""" Identify anatomy terms related to process/functions.
:param input_identifier: identifier for anatomy (including cell and cellular component)
"""
# we are looking for predicates that point either direction.
results = {'subject': [], 'object': []}
for direction, query in (('subject', ' ?anatID ?p $goID'),
('object', ' $goID ?p ?anatID')):
text = """
prefix UBERON: <http://purl.obolibrary.org/obo/UBERON_>
prefix GO: <http://purl.obolibrary.org/obo/GO_>
select distinct ?anatID ?anatLabel ?p ?pLabel
from <http://reasoner.renci.org/nonredundant>
from <http://reasoner.renci.org/ontology>
where {
graph <http://reasoner.renci.org/redundant> {
""" + query + """
}
graph <http://reasoner.renci.org/ontology/closure> {
?anatID rdfs:subClassOf UBERON:0001062 .
}
?anatID rdfs:label ?anatLabel .
?p rdfs:label ?pLabel
}
"""
results[direction] += self.triplestore.query_template(
inputs={'goID': input_identifier},
outputs=['anatID', 'anatLabel', 'p', 'pLabel'],
template_text=text)
return results
def pheno_or_disease_to_go(self, identifier):
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix BFO: <http://purl.obolibrary.org/obo/BFO_>
prefix GO: <http://purl.obolibrary.org/obo/GO_>
prefix MONDO: <http://purl.obolibrary.org/obo/MONDO_>
prefix HP: <http://purl.obolibrary.org/obo/MONDO_>
select distinct ?goID ?goLabel ?p ?pLabel
from <http://reasoner.renci.org/nonredundant>
from <http://reasoner.renci.org/ontology>
where {
graph <http://reasoner.renci.org/redundant> {
$input_id ?p ?goID .
}
graph <http://reasoner.renci.org/ontology/closure> {
{ ?goID rdfs:subClassOf GO:0008150 . }
UNION
{ ?goID rdfs:subClassOf GO:0003674 . }
}
?goID rdfs:label ?goLabel .
?p rdfs:label ?pLabel .
}
"""
results = self.triplestore.query_template(
inputs={'input_id': identifier},
outputs=['goID', 'goLabel', 'p', 'pLabel'],
template_text=text)
return results
def phenotype_to_anatomy(self, hp_identifier):
""" Identify anatomies related to phenotypes.
:param cell: HP identifier for phenotype
"""
#The subclassof uberon:0001062 ensures that the result
#is an anatomical entity.
#We don't need to do the subject/object game because there's nothing in ubergraph
# that goes that direction
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX UBERON: <http://purl.obolibrary.org/obo/UBERON_>
SELECT DISTINCT ?anatomy_id ?anatomy_label ?predicate ?predicate_label
FROM <http://reasoner.renci.org/ontology>
WHERE {
graph <http://reasoner.renci.org/redundant>{
$HPID ?predicate ?anatomy_id.
}
graph <http://reasoner.renci.org/ontology/closure>{
?anatomy_id rdfs:subClassOf UBERON:0001062.
}
?anatomy_id rdfs:label ?anatomy_label .
OPTIONAL {?predicate rdfs:label ?predicate_label.}
}
"""
results = self.triplestore.query_template(
inputs = { 'HPID': hp_identifier }, \
outputs = [ 'anatomy_id', 'anatomy_label', 'predicate', 'predicate_label'],\
template_text = text \
)
return results
def anatomy_to_phenotype(self, uberon_id):
#sparql very identical to phenotype_to_anatomy. could not find any anatomical
# entity that is a subject of subclass of HP:0000118, in ubergraph at this point.
# treating this as another version of pheno -> anatomical_entity but when
# anatomical_entity is known an
# we want to go back to a phenotype.
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX HP:<http://purl.obolibrary.org/obo/HP_>
SELECT DISTINCT ?pheno_id ?pheno_label ?predicate ?predicate_label
FROM <http://reasoner.renci.org/ontology>
WHERE {
graph <http://reasoner.renci.org/redundant> {
?pheno_id ?predicate $UBERONID.
}
graph <http://reasoner.renci.org/ontology/closure>{
?pheno_id rdfs:subClassOf HP:0000118.
}
?pheno_id rdfs:label ?pheno_label.
OPTIONAL {?predicate rdfs:label ?predicate_label.}
}"""
results = self.triplestore.query_template(
inputs = { 'UBERONID': uberon_id }, \
outputs = [ 'pheno_id', 'pheno_label', 'predicate', 'predicate_label' ],\
template_text = text \
)
return results
def biological_process_or_activity_to_chemical(self, go_id):
"""
Given a chemical finds associated GO Molecular Activities.
"""
results = []
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX GO: <http://purl.obolibrary.org/obo/GO_>
PREFIX RO: <http://purl.obolibrary.org/obo/RO_>
PREFIX chemical_entity: <http://purl.obolibrary.org/obo/CHEBI_24431>
PREFIX chemical_class: <http://purl.obolibrary.org/obo/CHEBI_24431>
SELECT DISTINCT ?chebi_id ?predicate ?label_predicate ?chebi_label
from <http://reasoner.renci.org/ontology>
from <http://reasoner.renci.org/nonredundant>
where {
$GO_ID ?predicate ?chebi_id.
?chebi_id rdfs:label ?chebi_label.
GRAPH <http://reasoner.renci.org/ontology/closure>
{ ?chebi_id rdfs:subClassOf chemical_class:.}
?predicate rdfs:label ?label_predicate.
FILTER ( datatype(?label_predicate) = xsd:string)
}
"""
results = self.triplestore.query_template(template_text=text,
outputs=[
'chebi_id', 'predicate',
'label_predicate',
'chebi_label'
],
inputs={'GO_ID': go_id})
return results
def pheno_to_biological_activity(self, pheno_id):
"""
Finds biological activities related to a phenotype
:param :pheno_id phenotype identifier
"""
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX owl: <http://www.w3.org/2002/07/owl#>
PREFIX GO: <http://purl.obolibrary.org/obo/GO_>
PREFIX has_phenotype_affecting: <http://purl.obolibrary.org/obo/UPHENO_0000001>
PREFIX RO: <http://purl.obolibrary.org/obo/RO_>
prefix HP: <http://purl.obolibrary.org/obo/HP_>
SELECT DISTINCT ?go_id ?predicate ?predicate_label ?go_label
from <http://reasoner.renci.org/nonredundant>
from <http://reasoner.renci.org/ontology>
WHERE {
$pheno_type ?predicate ?go_id.
?go_id rdfs:label ?go_label.
graph <http://reasoner.renci.org/ontology/closure> {
{ ?go_id rdfs:subClassOf GO:0008150 . }
UNION
{ ?go_id rdfs:subClassOf GO:0003674 . }
}
?predicate rdfs:label ?predicate_label.
}
"""
results = self.triplestore.query_template(
template_text=text,
inputs={'pheno_type': pheno_id},
outputs=['go_id', 'predicate', 'predicate_label', 'go_label'])
return results
def disease_to_anatomy(self, disease_id):
#THere are no anatomy-(predicate)->disease triples
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX anatomicalEntity: <http://purl.obolibrary.org/obo/UBERON_0001062>
SELECT DISTINCT ?anatomyID ?predicate ?predicate_label ?anatomy_label
FROM <http://reasoner.renci.org/nonredundant>
FROM <http://reasoner.renci.org/ontology>
WHERE {
graph <http://reasoner.renci.org/redundant> {
$diseaseID ?predicate ?anatomyID.
}
?anatomyID rdfs:label ?anatomy_label.
graph <http://reasoner.renci.org/ontology/closure> {
?anatomyID rdfs:subClassOf anatomicalEntity: .
}
?predicate rdfs:label ?predicate_label.
}
"""
results = []
results = self.triplestore.query_template(
template_text=text,
outputs=[
'anatomyID', 'predicate', 'predicate_label', 'anatomy_label'
],
inputs={'diseaseID': disease_id})
return results
def anatomy_to_chemical_substance(self, anatomy_id):
#There's no chemical-(predicate)->anatomy
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX chemical_entity: <http://purl.obolibrary.org/obo/CHEBI_24431>
SELECT DISTINCT ?predicate ?predicate_label ?chemical_entity ?chemical_label
FROM <http://reasoner.renci.org/ontology>
FROM <http://reasoner.renci.org/redundant>
WHERE {
$anatomy_id ?predicate ?chemical_entity.
graph <http://reasoner.renci.org/ontology/closure>
{
?chemical_entity rdfs:subClassOf chemical_entity:.
}
?predicate rdfs:label ?predicate_label .
?chemical_entity rdfs:label ?chemical_label.
}
"""
results = []
results = self.triplestore.query_template(
template_text=text,
outputs=[
'predicate', 'predicate_label', 'chemical_entity',
'chemical_label'
],
inputs={'anatomy_id': anatomy_id})
return results
def anatomy_to_disease(self, anatomy_id):
text = """
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX disease: <http://purl.obolibrary.org/obo/MONDO_0000001>
SELECT DISTINCT ?predicate ?predicate_label ?disease ?disease_label
FROM <http://reasoner.renci.org/ontology>
FROM <http://reasoner.renci.org/redundant>{
?disease ?predicate $anatomy_id.
graph <http://reasoner.renci.org/ontology/closure>
{
?disease rdfs:subClassOf disease:.
}
?predicate rdfs:label ?predicate_label .
?disease rdfs:label ?disease_label.
}
"""
results = []
results = self.triplestore.query_template(
template_text=text,
outputs=[
'predicate', 'predicate_label', 'disease', 'disease_label'
],
inputs={'anatomy_id': anatomy_id})
return results
def create_phenotype_anatomy_edge(self, node_id, node_label, input_id,
phenotype_node):
predicate = LabeledID(identifier='GAMMA:0000002',
label='inverse of has phenotype affecting')
anatomy_node = KNode(Text.obo_to_curie(node_id),
type=node_types.ANATOMICAL_ENTITY,
name=node_label)
edge = self.create_edge(anatomy_node, phenotype_node,
'uberongraph.get_anatomy_by_phenotype_graph',
input_id, predicate)
#node.name = node_label
return edge, anatomy_node
def create_anatomy_phenotype_edge(self, node_id, node_label, input_id,
anatomy_node):
predicate = LabeledID(identifier='GAMMA:0000002',
label='inverse of has phenotype affecting')
phenotype_node = KNode(Text.obo_to_curie(node_id),
type=node_types.PHENOTYPIC_FEATURE,
name=node_label)
edge = self.create_edge(anatomy_node, phenotype_node,
'uberongraph.get_phenotype_by_anatomy_graph',
input_id, predicate)
#node.name = node_label
return edge, phenotype_node
def dep_get_anatomy_by_phenotype_graph(self, phenotype_node):
results = []
for curie in phenotype_node.get_synonyms_by_prefix('HP'):
anatomies = self.phenotype_to_anatomy(curie)
for r in anatomies:
node = KNode(r['anatomy_id'],
type=node_types.ANATOMICAL_ENTITY,
name=r['anatomy_label'])
# try to derive the label from the relation for the new ubergraph axioms
predicate_label = r['predicate_label'] or '_'.join(
r['predicate'].split('#')[-1].split('.'))
predicate = LabeledID(Text.obo_to_curie(r['predicate']),
predicate_label)
edge = self.create_edge(
phenotype_node, node,
'uberongraph.get_anatomy_by_phenotype_graph',
phenotype_node.id, predicate)
# edge, node = self.create_phenotype_anatomy_edge(r['anatomy_id'],r['anatomy_label'],curie,phenotype_node)
if phenotype_node.name is None:
phenotype_node.name = r['input_label']
results.append((edge, node))
#These tend to be very high level terms. Let's also get their parts to
#be more inclusive.
#TODO: there ought to be a more principled way to take care of this, but
#it highlights the uneasy relationship between the high level world of
#smartapi and the low-level sparql-vision.
part_results = self.get_anatomy_parts(r['anatomy_id'])
for pr in part_results:
# pedge, pnode = self.create_phenotype_anatomy_edge(pr['part'],pr['partlabel'],curie,phenotype_node)
pnode = KNode(pr['part'],
type=node_types.ANATOMICAL_ENTITY,
name=pr['partlabel'])
pedge = self.create_edge(
phenotype_node, pnode,
'uberongraph.get_anatomy_by_phenotype_graph',
phenotype_node.id, predicate)
results.append((pedge, pnode))
return results
def get_out_by_in(self,
input_node,
output_type,
prefixes,
subject=True,
object=True):
returnresults = []
caller = f'uberongraph.{inspect.stack()[1][3]}'
results = {'subject': [], 'object': []}
curies = set()
for pre in prefixes:
curies.update(input_node.get_synonyms_by_prefix(pre))
for curie in curies:
results['subject'] += self.get_neighbor(curie,
output_type,
subject=True)
results['object'] += self.get_neighbor(curie,
output_type,
subject=False)
for direction in ['subject', 'object']:
done = set()
for r in results[direction]:
key = (r['p'], r['output_id'])
if key in done:
continue
predicate_curie = Text.obo_to_curie(r['p'])
prefix = Text.get_curie(predicate_curie)
prefix = prefix if prefix == 'ubergraph-axioms.ofn' else prefix.upper(
)
upper_cased_predicate_curie = prefix + ":" + Text.un_curie(
predicate_curie)
predicate = LabeledID(upper_cased_predicate_curie, r['pLabel'])
output_node = KNode(r['output_id'],
type=output_type,
name=r['output_label'])
if direction == 'subject':
edge = self.create_edge(input_node, output_node, caller,
curie, predicate)
else:
edge = self.create_edge(output_node, input_node, caller,
curie, predicate)
done.add(key)
returnresults.append((edge, output_node))
return returnresults
#Don't get confused. There is the direction of the statement (who is the subject
# and who is the object) and which of them we are querying by. We want to query
# independent of direction i.e. let the input node be either the subject or the object.
def get_anatomy_by_anatomy_graph(self, anatomy_node):
return self.get_out_by_in(anatomy_node, node_types.ANATOMICAL_ENTITY,
['UBERON', 'CL', 'GO'])
def get_phenotype_by_anatomy_graph(self, anatomy_node):
return self.get_out_by_in(anatomy_node, node_types.PHENOTYPIC_FEATURE,
['UBERON', 'CL', 'GO'])
def get_chemical_substance_by_anatomy(self, anatomy_node):
return self.get_out_by_in(anatomy_node, node_types.CHEMICAL_SUBSTANCE,
['UBERON', 'CL', 'GO'])
def get_process_by_anatomy(self, anatomy_node):
return self.get_out_by_in(anatomy_node, node_types.BIOLOGICAL_PROCESS,
['UBERON', 'CL', 'GO'])
def get_activity_by_anatomy(self, anatomy_node):
return self.get_out_by_in(anatomy_node, node_types.MOLECULAR_ACTIVITY,
['UBERON', 'CL', 'GO'])
def get_disease_by_anatomy_graph(self, anatomy_node):
return self.get_out_by_in(anatomy_node, node_types.DISEASE,
['UBERON', 'CL', 'GO'])
def get_anatomy_by_process_or_activity(self, go_node):
return self.get_out_by_in(go_node, node_types.ANATOMICAL_ENTITY,
['GO'])
def get_chemical_entity_by_process_or_activity(self, go_node):
return self.get_out_by_in(go_node, node_types.CHEMICAL_SUBSTANCE,
['GO'])
def get_process_by_disease(self, disease_node):
return self.get_out_by_in(disease_node, node_types.BIOLOGICAL_PROCESS,
['MONDO'])
def get_activity_by_disease(self, disease_node):
return self.get_out_by_in(disease_node, node_types.MOLECULAR_ACTIVITY,
['MONDO'])
def get_anatomy_by_disease(self, disease_node):
return self.get_out_by_in(disease_node, node_types.ANATOMICAL_ENTITY,
['MONDO'])
def get_chemical_by_disease(self, disease_node):
return self.get_out_by_in(disease_node, node_types.CHEMICAL_SUBSTANCE,
['MONDO'])
def get_process_by_phenotype(self, pheno_node):
return self.get_out_by_in(pheno_node, node_types.BIOLOGICAL_PROCESS,
['HP'])
def get_chemical_by_phenotype(self, pheno_node):
return self.get_out_by_in(pheno_node, node_types.CHEMICAL_SUBSTANCE,
['HP'])
def get_activity_by_phenotype(self, pheno_node):
return self.get_out_by_in(pheno_node, node_types.MOLECULAR_ACTIVITY,
['HP'])
def get_anatomy_by_phenotype_graph(self, pheno_node):
return self.get_out_by_in(pheno_node, node_types.ANATOMICAL_ENTITY,
['HP'])
def get_chemical_by_chemical(self, chem_node):
return self.get_out_by_in(chem_node, node_types.CHEMICAL_SUBSTANCE,
['CHEBI'])
def disease_get_ancestors(self, disease_node):
curie = disease_node.id
prefix = Text.get_curie(curie)
if "MONDO" != prefix:
return []
query = f"""
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix MONDO: <http://purl.obolibrary.org/obo/MONDO_>
select distinct ?output_id ?label
where {{
graph <http://reasoner.renci.org/ontology/closure> {{
$disease_id rdfs:subClassOf ?output_id .
?output_id rdfs:subClassOf MONDO:0000001 .
}}
graph <http://reasoner.renci.org/ontology>{{
?output_id rdfs:label ?label.
}}
}}
"""
results = self.triplestore.query_template(
template_text=query,
inputs={'disease_id': curie},
outputs=['output_id', 'label'])
outputs = []
for row in results:
ancestor_node = KNode(
row['output_id'],
label=row['label'],
type=node_types.DISEASE_OR_PHENOTYPIC_FEATURE)
if ancestor_node.id == disease_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=disease_node,
target_node=ancestor_node,
predicate=predicate,
provided_by='uberongraph.disease_get_ancestors',
input_id=disease_node.id)
outputs.append((edge, ancestor_node))
return outputs
def __init__(self, context): #triplestore):
super(UberonGraphKS, self).__init__("uberongraph", context)
self.triplestore = TripleStore(self.url)