def format_edge(subject_id: str, object_id: str, predicate_label: str):
relation_curie = kg2_util.predicate_label_to_curie(
predicate_label, REACTOME_RELATION_CURIE_PREFIX)
if predicate_label == kg2_util.EDGE_LABEL_BIOLINK_SAME_AS:
return kg2_util.make_edge_biolink(subject_id, object_id,
predicate_label,
REACTOME_KB_CURIE_ID, None)
return kg2_util.make_edge(subject_id, object_id, relation_curie,
predicate_label, REACTOME_KB_CURIE_ID)
def format_edge(subject_id, object_id, predicate, update_date):
relation_curie = kg2_util.predicate_label_to_curie(
predicate, DRUGCENTRAL_RELATION_CURIE_PREFIX)
if predicate == kg2_util.EDGE_LABEL_BIOLINK_SAME_AS:
return kg2_util.make_edge_biolink(subject_id, object_id, predicate,
DRUGCENTRAL_SOURCE, update_date)
else:
return kg2_util.make_edge(subject_id, object_id, relation_curie,
predicate, DRUGCENTRAL_SOURCE, update_date)
def format_edge(subject_id: str,
object_id: str,
predicate_label: str,
description: str,
publications: list = None):
relation_curie = kg2_util.predicate_label_to_curie(
predicate_label, DRUGBANK_RELATION_CURIE_PREFIX)
edge = kg2_util.make_edge(subject_id, object_id, relation_curie,
predicate_label, DRUGBANK_KB_CURIE_ID, None)
if description is not None:
edge["publications_info"] = {"sentence": description}
if publications is not None:
edge["publications"] = publications
return edge
def make_hmdb_edge(subject_id: str, object_id: str, subject_prefix: str,
object_prefix: str, predicate_label: str, update_date: str,
publications_info: dict):
relation_curie = kg2_util.predicate_label_to_curie(predicate_label,
CURIE_PREFIX_HMDB)
subject = subject_prefix + ":" + subject_id
object = object_id
if object_prefix is not None:
object = object_prefix + ":" + object_id
if predicate_label == kg2_util.EDGE_LABEL_BIOLINK_SAME_AS:
edge = kg2_util.make_edge_biolink(subject, object, predicate_label,
HMDB_PROVIDED_BY_CURIE_ID,
update_date)
else:
edge = kg2_util.make_edge(subject, object, relation_curie,
predicate_label, HMDB_PROVIDED_BY_CURIE_ID,
update_date)
edge["publications_info"] = publications_info
return edge
kg2_util.merge_two_dicts(
{
'subject': result_item_list[0],
'object': result_item_list[2]
}, result_item_list[1]['data'])
for result_item_list in edges_result
]
for edge_dict in edges_list:
del edge_dict['is_defined_by']
del edge_dict['seed_node_uuid']
del edge_dict['source_node_uuid']
del edge_dict['target_node_uuid']
predicate_label = edge_dict['relation']
edge_dict['edge_label'] = predicate_label
del edge_dict['relation']
relation_curie = kg2_util.predicate_label_to_curie(
predicate_label, KG1_RELATION_CURIE_PREFIX)
if relation_curie == 'bioLink:subclass_of':
relation_curie = kg2_util.CURIE_ID_RDFS_SUBCLASS_OF
edge_dict['relation'] = relation_curie
edge_dict['negated'] = False
publications = edge_dict.get('publications', None)
if publications is not None and publications != '':
publications = publications.split(',')
else:
publications = []
edge_dict['publications'] = publications
edge_dict['update_date'] = None
provided_by = edge_dict['provided_by']
if provided_by.startswith('DGIdb;'):
provided_by = 'DGIdb'
provided_by_kg2 = KG1_PROVIDED_BY_TO_KG2_PROVIDED_BY_CURIE_IDS.get(
return [{"nodes": nodes, "edges": edges}, update_date]
if __name__ == '__main__':
print("Starting Script:", date())
args = get_args()
print("Starting PubMed ID Load:", date())
pmids = set(json.load(open(args.kg2PMIDs)))
print("Finishing PubMedID Load:", date(), ",", len(pmids), "PMIDs in KG2")
pubmed_dir = args.inputDirectory
nodes = []
edges = []
latest_date = 0
mesh_predicate_label = "references"
mesh_relation_curie = kg2_util.predicate_label_to_curie(
mesh_predicate_label, kg2_util.CURIE_PREFIX_PMID)
for filename in os.listdir(pubmed_dir):
if ".gz" in filename:
print("Starting Load of", filename, ":", date())
xml_file = gzip.open(pubmed_dir + filename)
data = xmltodict.parse(xml_file.read())
print("Finished Load of", filename, ":", date())
articles = data["PubmedArticleSet"]["PubmedArticle"]
for article in articles:
[data,
update_date] = make_node_and_edges(article,
mesh_predicate_label,
mesh_relation_curie)
for node in data["nodes"]:
nodes.append(node)