def process_belscript(file_name, **kwargs):
"""Return a PybelProcessor by processing a BEL script file.
Key word arguments are passed directly to pybel.from_path,
for further information, see
pybel.readthedocs.io/en/latest/io.html#pybel.from_path
Some keyword arguments we use here differ from the defaults
of PyBEL, namely we set `citation_clearing` to False
and `no_identifier_validation` to True.
Parameters
----------
file_name : str
The path to a BEL script file.
Returns
-------
bp : PybelProcessor
A PybelProcessor object which contains INDRA Statements in
bp.statements.
"""
if 'citation_clearing' not in kwargs:
kwargs['citation_clearing'] = False
if 'no_identifier_validation' not in kwargs:
kwargs['no_identifier_validation'] = True
pybel_graph = pybel.from_path(file_name, **kwargs)
return process_pybel_graph(pybel_graph)
def process_belscript(file_name, **kwargs):
"""Return a PybelProcessor by processing a BEL script file.
Key word arguments are passed directly to pybel.from_path,
for further information, see
pybel.readthedocs.io/en/latest/io.html#pybel.from_path
Some keyword arguments we use here differ from the defaults
of PyBEL, namely we set `citation_clearing` to False
and `no_identifier_validation` to True.
Parameters
----------
file_name : str
The path to a BEL script file.
Returns
-------
bp : PybelProcessor
A PybelProcessor object which contains INDRA Statements in
bp.statements.
"""
if 'citation_clearing' not in kwargs:
kwargs['citation_clearing'] = False
if 'no_identifier_validation' not in kwargs:
kwargs['no_identifier_validation'] = True
pybel_graph = pybel.from_path(file_name, **kwargs)
return process_pybel_graph(pybel_graph)
def setUpClass(cls):
"""Set up this class with several pre-loaded BEL graphs."""
super(TestInterchange, cls).setUpClass()
with mock_bel_resources:
cls.thorough_graph = from_path(test_bel_thorough,
manager=cls.manager,
allow_nested=True)
cls.slushy_graph = from_path(
test_bel_slushy,
manager=cls.manager,
disallow_unqualified_translocations=True)
cls.simple_graph = from_url(Path(test_bel_simple).as_uri(),
manager=cls.manager)
cls.isolated_graph = from_path(test_bel_isolated,
manager=cls.manager)
cls.misordered_graph = from_path(test_bel_misordered,
manager=cls.manager,
citation_clearing=False)
def bel_graph_loader(from_dir: str) -> BELGraph:
"""Obtains a combined BELGraph from all the BEL documents in one folder.
:param from_dir: The folder with the BEL documents.
:return: A corresponding BEL Graph.
"""
logger.info("Loading BEL Graph.")
files = [
join(from_dir, file) for file in listdir(from_dir)
if isfile(join(from_dir, file))
]
bel_files = [file for file in files if file[-4:].lower() == '.bel']
bel_graphs = [from_path(file) for file in bel_files]
return union(bel_graphs)
def load_paths(paths, connection=None):
"""Parses multiple BEL scripts with :func:`pybel.from_path` and returns the union of the resulting graphs.
:param iter[str] paths: An iterable over paths to BEL scripts
:param connection: A custom database connection string or manager
:type connection: Optional[str or pybel.manager.Manager]
:rtype: pybel.BELGraph
"""
manager = Manager.ensure(connection)
return union(
from_path(path, manager=manager)
for path in paths
)
def process_belscript(file_name, **kwargs):
"""Return a PybelProcessor by processing a BEL script file.
Key word arguments are passed directly to pybel.from_path,
for further information, see
pybel.readthedocs.io/en/latest/io.html#pybel.from_path
Parameters
----------
file_name : str
The path to a BEL script file.
Returns
-------
bp : PybelProcessor
A PybelProcessor object which contains INDRA Statements in
bp.statements.
"""
pybel_graph = pybel.from_path(file_name, **kwargs)
return process_pybel_graph(pybel_graph)
def test_from_path(self, m1, m2, m3):
graph = from_path(test_bel_extensions, manager=self.manager)
self.assertEqual(0, len(graph.warnings))
self.assertEqual(expected_test_bel_4_metadata, graph.document)
self.assertEqual(
{
'WINE':
wine_iri,
'PIZZA':
'http://www.lesfleursdunormal.fr/static/_downloads/pizza_onto.owl'
}, graph.namespace_owl)
self.assertEqual({'Wine': wine_iri}, graph.annotation_owl)
self.assertEqual({'HGNC': HGNC_URL}, graph.namespace_url)
a = PROTEIN, 'HGNC', 'AKT1'
b = PROTEIN, 'HGNC', 'EGFR'
self.assertHasNode(graph, a)
self.assertHasNode(graph, b)
self.assertHasEdge(graph, a, b)
annots = {
CITATION: {
CITATION_NAME: 'That one article from last week',
CITATION_REFERENCE: '123455',
CITATION_TYPE: 'PubMed'
},
EVIDENCE: 'Made up support, not even qualifying as evidence',
ANNOTATIONS: {
'Wine': {
'Cotturi': True
}
}
}
self.assertHasEdge(graph, (ABUNDANCE, "PIZZA", "MeatTopping"),
(ABUNDANCE, 'WINE', 'Wine'), **annots)
self.assertHasEdge(graph, (ABUNDANCE, "PIZZA", "TomatoTopping"),
(ABUNDANCE, 'WINE', 'Wine'), **annots)
self.assertHasEdge(graph, (ABUNDANCE, 'WINE', 'WhiteWine'),
(ABUNDANCE, "PIZZA", "FishTopping"), **annots)
def load_paths(paths, connection=None):
"""Loads a group of BEL graphs.
Internally, this function uses a shared :class:`pybel.parser.MetadataParser` to cache the definitions more
efficiently.
:param paths: An iterable over paths to BEL scripts
:param paths: iter
:param connection: A custom database connection string
:type connection: str
:return: A BEL graph comprised of the union of all BEL graphs produced by each BEL script
:rtype: pybel.BELGraph
"""
metadata_parser = build_metadata_parser(connection)
result = BELGraph()
for path in paths:
subgraph = from_path(path, manager=metadata_parser)
left_merge(result, subgraph)
return result
def setUp(self):
if 'PYBEL_BASE' in os.environ:
test_bel_simple_path = os.path.join(os.environ['PYBEL_BASE'], 'tests', 'bel', 'test_bel.bel')
self.graph = pybel.from_path(test_bel_simple_path)
else:
test_bel_simple_url = 'https://raw.githubusercontent.com/pybel/pybel/develop/tests/bel/test_bel.bel'
self.graph = pybel.from_url(test_bel_simple_url)
infer_central_dogma(self.graph)
n1 = GENE, 'HGNC', 'AKT1'
n2 = RNA, 'HGNC', 'EGFR'
n3 = GENE, 'HGNC', 'DUMMY1'
self.graph.add_simple_node(*n3)
n4 = GENE, 'HGNC', 'DUMMY2'
self.graph.add_simple_node(*n4)
self.graph.add_edge(n1, n3)
self.graph.add_edge(n2, n4)
def convert_recursive(directory, connection=None, upload=False, pickle=False, store_parts=False,
enrich_citations=False):
"""Recursively parses and either uploads/pickles graphs in a given directory and sub-directories"""
metadata_parser = build_metadata_parser(connection)
paths = list(get_paths_recursive(directory))
log.info('Paths to parse: %s', paths)
for path in paths:
try:
graph = from_path(path, manager=metadata_parser.manager)
except:
log.exception('Problem parsing %s', path)
continue
if enrich_citations:
fix_pubmed_citations(graph)
if upload:
safe_upload(metadata_parser.manager, graph, store_parts=store_parts)
if pickle:
new_path = '{}.gpickle'.format(path[:-4]) # [:-4] gets rid of .bel at the end of the file name
to_pickle(graph, new_path)
import sys
import pickle
import pybel
from pybel.struct.filters import has_protein_modification
from indra.sources import bel
from indra.sources.bel.processor import get_agent
from .util import get_mod_sites
if __name__ == '__main__':
# Parse the BEL script, takes a few minutes
if sys.argv[1] == 'parse_belscript':
input_file = sys.argv[2]
output_file = sys.argv[3]
pbg = pybel.from_path(input_file)
pybel.to_pickle(pbg, output_file)
# Get all variant sites from the graph
#elif sys.argv[1] == 'get_pybel_mod_agents':
# pbg = pybel.from_pickle('output/large_corpus_pybel.pkl')
# mod_nodes = [get_agent(n) for n in pbg.nodes()
# if has_protein_modification(n)]
# with open('output/bel_mod_agents.pkl', 'wb') as f:
# pickle.dump(mod_nodes, f)
elif sys.argv[1] == 'get_pybel_stmts_by_site':
input_file = sys.argv[2]
output_file = sys.argv[3]
pbg = pybel.from_pickle(input_file)
pbp = bel.process_pybel_graph(pbg)
sites = get_mod_sites(pbp.statements)
with open(output_file, 'wb') as f:
pickle.dump(sites, f)
else:
def get_misordered_graph(mock_get):
return from_path(test_bel_misordered,
manager=cls.manager,
citation_clearing=False)
def get_isolated_graph(mock_get):
return from_path(test_bel_isolated, manager=cls.manager)
def get_slushy_graph(mock):
return from_path(test_bel_slushy, manager=cls.manager)
def get_thorough_graph(mock):
return from_path(test_bel_thorough,
manager=cls.manager,
allow_nested=True)
def convert_paths(paths,
connection=None,
upload=False,
pickle=False,
canonicalize=True,
infer_central_dogma=True,
enrich_citations=False,
send=False,
version_in_path=False,
**kwargs):
"""Recursively parses and either uploads/pickles graphs in a given set of files
:param iter[str] paths: The paths to convert
:param connection: The connection
:type connection: None or str or pybel.manager.Manager
:param bool upload: Should the networks be uploaded to the cache?
:param bool pickle: Should the networks be saved as pickles?
:param bool canonicalize: Calculate canonical nodes?
:param bool infer_central_dogma: Should the central dogma be inferred for all proteins, RNAs, and miRNAs
:param bool enrich_citations: Should the citations be enriched using Entrez Utils?
:param bool send: Send to PyBEL Web?
:param bool version_in_path: Add the current pybel version to the pathname
:param kwargs: Parameters to pass to :func:`pybel.from_path`
"""
manager = Manager.ensure(connection)
failures = []
for path in paths:
log.info('parsing: %s', path)
try:
graph = from_path(path, manager=manager, **kwargs)
except Exception as e:
log.exception('problem parsing %s', path)
failures.append((path, e))
continue
if canonicalize:
add_canonical_names(graph)
if infer_central_dogma:
infer_central_dogma_mutator(graph)
if enrich_citations:
enrich_pubmed_citations(graph=graph, manager=manager)
if upload:
to_database(graph, connection=manager, store_parts=True)
if pickle:
name = path[:-len(
'.bel')] # gets rid of .bel at the end of the file name
if version_in_path:
new_path = '{}-{}.gpickle'.format(name, get_pybel_version())
else:
new_path = '{}.gpickle'.format(name)
to_pickle(graph, new_path)
log.info('output pickle: %s', new_path)
if send:
response = to_web(graph)
log.info('sent to PyBEL Web with response: %s', response.json())
return failures
