def test_update_matrix_pmods(self):
"""Test updating the matrix with multiple protein modifications."""
sub = protein(namespace='HGNC', name='A', identifier='1')
obj = protein(namespace='HGNC',
name='B',
identifier='2',
variants=[pmod('Ub'), pmod('Ph')])
index = {'A', 'B'}
test_dict = {}
test_matrix = DataFrame(0, index=index, columns=index)
test_dict["activation_ubiquination"] = test_matrix
test_dict["activation_phosphorylation"] = test_matrix
update_spia_matrices(test_dict, sub, obj, {'relation': 'increases'})
self.assertEqual(test_dict["activation_ubiquination"]['A']['B'], 1)
self.assertEqual(test_dict["activation_ubiquination"]['A']['A'], 0)
self.assertEqual(test_dict["activation_ubiquination"]['B']['A'], 0)
self.assertEqual(test_dict["activation_ubiquination"]['B']['B'], 0)
self.assertEqual(test_dict["activation_phosphorylation"]['A']['B'], 1)
self.assertEqual(test_dict["activation_phosphorylation"]['A']['A'], 0)
self.assertEqual(test_dict["activation_phosphorylation"]['B']['A'], 0)
self.assertEqual(test_dict["activation_phosphorylation"]['B']['B'], 0)
def test_canonicalize_variant_dsl(self):
"""Use the __str__ functions in the DSL to create BEL instead of external pybel.canonicalize."""
self.assertEqual('var("p.Val600Glu")', str(hgvs('p.Val600Glu')))
self.assertEqual('var("p.Val600Glu")',
str(protein_substitution('Val', 600, 'Glu')))
self.assertEqual('pmod(Ph)', str(pmod('Ph')))
self.assertEqual('pmod(TEST:Ph)', str(pmod('Ph', namespace='TEST')))
self.assertEqual('pmod(TEST:Ph, Ser)',
str(pmod('Ph', namespace='TEST', code='Ser')))
self.assertEqual(
'pmod(TEST:Ph, Ser, 5)',
str(pmod('Ph', namespace='TEST', code='Ser', position=5)))
self.assertEqual(
'pmod(GO:"protein phosphorylation", Thr, 308)',
str(
pmod(name='protein phosphorylation',
namespace='GO',
code='Thr',
position=308)))
self.assertEqual('frag("?")', str(fragment()))
self.assertEqual('frag("672_713")', str(fragment(start=672, stop=713)))
self.assertEqual('frag("?", "descr")',
str(fragment(description='descr')))
self.assertEqual(
'frag("672_713", "descr")',
str(fragment(start=672, stop=713, description='descr')))
self.assertEqual('gmod(Me)', str(gmod('Me')))
self.assertEqual('gmod(TEST:Me)', str(gmod('Me', namespace='TEST')))
self.assertEqual('gmod(GO:"DNA Methylation")',
str(gmod('DNA Methylation', namespace='GO')))
def _help_test_negative_correlation_with_object_variant(self, x):
statement = f'kin(p(FPLX:GSK3)) neg p(HGNC:MAPT,pmod({x}))'
result = self.parser.relation.parseString(statement)
expected_dict = {
SOURCE: {
MODIFIER: ACTIVITY,
EFFECT: activity_mapping['kin'],
FUNCTION: PROTEIN,
CONCEPT: {
NAMESPACE: 'FPLX',
NAME: 'GSK3',
},
},
RELATION: NEGATIVE_CORRELATION,
TARGET: {
FUNCTION: PROTEIN,
CONCEPT: {
NAMESPACE: 'HGNC',
NAME: 'MAPT',
},
VARIANTS: [pmod('Ph')]
}
}
self.assertEqual(expected_dict, result.asDict())
sub = protein('FPLX', 'GSK3')
self.assert_has_node(sub)
obj = protein('HGNC', 'MAPT', variants=pmod('Ph'))
self.assert_has_node(obj)
self.assert_has_edge(sub, obj, relation=expected_dict[RELATION])
self.assert_has_edge(obj, sub, relation=expected_dict[RELATION])
def test_positiveCorrelation_withSelfReferential(self):
"""
3.2.2 http://openbel.org/language/web/version_2.0/bel_specification_version_2.0.html#XposCor
Self-referential relationships"""
statement = 'p(HGNC:GSK3B, pmod(P, S, 9)) pos act(p(HGNC:GSK3B), ma(kin))'
result = self.parser.relation.parseString(statement)
expected_dict = {
SUBJECT: {
FUNCTION: PROTEIN,
CONCEPT: {
NAMESPACE: 'HGNC',
NAME: 'GSK3B',
},
VARIANTS: [pmod('Ph', position=9, code='Ser')]
},
RELATION: POSITIVE_CORRELATION,
OBJECT: {
MODIFIER: ACTIVITY,
TARGET: {
FUNCTION: PROTEIN,
CONCEPT: {
NAMESPACE: 'HGNC',
NAME: 'GSK3B',
},
},
EFFECT: {
NAME: 'kin',
NAMESPACE: BEL_DEFAULT_NAMESPACE
}
},
}
self.assertEqual(expected_dict, result.asDict())
subject_node = protein('HGNC',
'GSK3B',
variants=pmod('Ph', code='Ser', position=9))
self.assert_has_node(subject_node)
object_node = protein('HGNC', 'GSK3B')
self.assert_has_node(object_node)
self.assert_has_edge(subject_node,
object_node,
relation=expected_dict[RELATION])
self.assert_has_edge(object_node,
subject_node,
relation=expected_dict[RELATION])
def test_transphosphorylation():
egfr = Agent('EGFR', db_refs={'HGNC': id('EGFR')})
egfr_dimer = Agent('EGFR',
bound_conditions=[BoundCondition(egfr)],
db_refs={'HGNC': id('EGFR')})
stmt = Transphosphorylation(egfr_dimer, 'Y', '1173')
stmt_hash = stmt.get_hash(refresh=True)
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert belgraph.number_of_nodes() == 3
assert belgraph.number_of_edges() == 3
egfr_dimer_node = complex_abundance([egfr_dsl, egfr_dsl])
egfr_phos_node = egfr_dsl.with_variants(pmod('Ph', 'Tyr', 1173))
edge_data = get_edge_data(belgraph, egfr_dimer_node, egfr_phos_node)
assert edge_data == {
pc.RELATION: pc.DIRECTLY_INCREASES,
pc.ANNOTATIONS: {
'stmt_hash': {
stmt_hash: True
},
'uuid': {
stmt.uuid: True
},
'belief': {
stmt.belief: True
},
},
}, edge_data
def test_update_matrix_inhibition_ubiquination(self):
"""Test updating the matrix with an inhibition ubiquitination."""
sub = protein(namespace='HGNC', name='A', identifier='1')
obj = protein(namespace='HGNC',
name='B',
identifier='2',
variants=[pmod('Ub')])
index = {'A', 'B'}
test_dict = {}
test_matrix = DataFrame(0, index=index, columns=index)
# Initialize matrix correctly
self.assertEqual(test_matrix.values.all(), 0)
test_dict["inhibition_ubiquination"] = test_matrix
update_spia_matrices(test_dict, sub, obj, {'relation': 'decreases'})
self.assertEqual(test_dict["inhibition_ubiquination"]['A']['B'], 1)
self.assertEqual(test_dict["inhibition_ubiquination"]['A']['A'], 0)
self.assertEqual(test_dict["inhibition_ubiquination"]['B']['A'], 0)
self.assertEqual(test_dict["inhibition_ubiquination"]['B']['B'], 0)
def test_gef():
gef = Agent('SOS1',
mods=[ModCondition('phosphorylation')],
db_refs={'HGNC': '11187'})
ras = Agent('KRAS', db_refs={'HGNC': '6407'})
stmt = Gef(gef, ras)
stmt_hash = stmt.get_hash(refresh=True)
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert len(belgraph) == 3
assert belgraph.number_of_edges() == 2
gef_reference_node = protein(namespace='HGNC',
name='SOS1',
identifier='11187')
gef_node = gef_reference_node.with_variants(pmod('Ph'))
assert gef_reference_node in belgraph
assert gef_node in belgraph
assert kras_node in belgraph
edge_data = get_edge_data(belgraph, gef_node, kras_node)
edge = {
pc.RELATION: pc.DIRECTLY_INCREASES,
pc.SUBJECT: activity('gef'),
pc.OBJECT: activity('gtp'),
pc.ANNOTATIONS: {
'stmt_hash': stmt_hash,
'uuid': stmt.uuid,
'belief': stmt.belief,
},
}
assert edge_data == edge, edge_data
def test_complex_with_complex():
grb2 = Agent('GRB2', db_refs={'HGNC': id('GRB2')})
egfr_grb2 = Agent('EGFR', db_refs={'HGNC': id('EGFR')},
bound_conditions=[BoundCondition(grb2)])
sos1_phos = Agent('SOS1',
mods=[ModCondition('phosphorylation', 'Y', '100')],
db_refs={'HGNC': id('SOS1')})
stmt = Complex([sos1_phos, egfr_grb2])
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert len(belgraph) == 6
assert belgraph.number_of_edges() == 5
egfr_grb2_complex = complex_abundance([egfr_dsl, grb2_dsl])
egfr_grb2_complex_sos1_phos_complex = complex_abundance([
egfr_grb2_complex,
sos1_dsl.with_variants(pmod('Ph', 'Tyr', 100))
])
assert egfr_grb2_complex in belgraph
for member in egfr_grb2_complex.members:
assert member in belgraph
assert egfr_grb2_complex_sos1_phos_complex in belgraph
for member in egfr_grb2_complex_sos1_phos_complex.members:
assert member in belgraph
def test_gap():
gap = Agent('RASA1', mods=[ModCondition('phosphorylation')],
db_refs={'HGNC': '9871'})
ras = Agent('KRAS', db_refs={'HGNC': '6407'})
stmt = Gap(gap, ras)
stmt_hash = stmt.get_hash(refresh=True)
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert len(belgraph) == 3
assert belgraph.number_of_edges() == 2
gap_reference_node = protein(
namespace='HGNC', name='RASA1', identifier='9871')
gap_node = gap_reference_node.with_variants(pmod('Ph'))
ras_node = protein(namespace='HGNC', name='KRAS', identifier='6407')
assert gap_reference_node in belgraph
assert gap_node in belgraph
assert ras_node in belgraph
edge_data = get_edge_data(belgraph, gap_node, ras_node)
edge = {
pc.RELATION: pc.DIRECTLY_DECREASES,
pc.SUBJECT: activity('gap'),
pc.OBJECT: activity('gtp'),
pc.ANNOTATIONS: {
'stmt_hash': {stmt_hash: True},
'uuid': {stmt.uuid: True},
'belief': {stmt.belief: True},
},
}
assert edge_data == edge, edge_data
def test_complex_with_complex():
grb2 = Agent('GRB2', db_refs={'HGNC': id('GRB2')})
egfr_grb2 = Agent('EGFR', db_refs={'HGNC': id('EGFR')},
bound_conditions=[BoundCondition(grb2)])
sos1_phos = Agent('SOS1',
mods=[ModCondition('phosphorylation', 'Y', '100')],
db_refs={'HGNC': id('SOS1')})
stmt = Complex([sos1_phos, egfr_grb2])
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert len(belgraph) == 6
assert belgraph.number_of_edges() == 5
egfr_grb2_complex = complex_abundance([egfr_dsl, grb2_dsl])
egfr_grb2_complex_sos1_phos_complex = complex_abundance([
egfr_grb2_complex,
sos1_dsl.with_variants(pmod('Ph', 'Tyr', 100))
])
assert egfr_grb2_complex in belgraph
for member in egfr_grb2_complex.members:
assert member in belgraph
assert egfr_grb2_complex_sos1_phos_complex in belgraph
for member in egfr_grb2_complex_sos1_phos_complex.members:
assert member in belgraph
def as_bel(self) -> protein:
parent = self.protein.as_bel()
variant = pmod(
name=self.modification_type.name,
position=self.position,
code=amino_acid_dict[self.residue.upper()],
)
return parent.with_variants(variant)
def test_negativeCorrelation_withObjectVariant(self):
"""
3.2.1 http://openbel.org/language/web/version_2.0/bel_specification_version_2.0.html#XnegCor
Test phosphoralation tag"""
statement = 'kin(p(FPLX:GSK3)) neg p(HGNC:MAPT,pmod(P))'
result = self.parser.relation.parseString(statement)
expected_dict = {
SUBJECT: {
MODIFIER: ACTIVITY,
EFFECT: {
NAME: 'kin',
NAMESPACE: BEL_DEFAULT_NAMESPACE
},
TARGET: {
FUNCTION: PROTEIN,
CONCEPT: {
NAMESPACE: 'FPLX',
NAME: 'GSK3',
},
},
},
RELATION: NEGATIVE_CORRELATION,
OBJECT: {
FUNCTION: PROTEIN,
CONCEPT: {
NAMESPACE: 'HGNC',
NAME: 'MAPT',
},
VARIANTS: [pmod('Ph')]
}
}
self.assertEqual(expected_dict, result.asDict())
sub = protein('FPLX', 'GSK3')
self.assert_has_node(sub)
obj = protein('HGNC', 'MAPT', variants=pmod('Ph'))
self.assert_has_node(obj)
self.assert_has_edge(sub, obj, relation=expected_dict[RELATION])
self.assert_has_edge(obj, sub, relation=expected_dict[RELATION])
def _help_test_pmod_simple(self, statement):
result = self.parser.parseString(statement)
expected = {
KIND: PMOD,
IDENTIFIER: {
NAMESPACE: BEL_DEFAULT_NAMESPACE,
NAME: 'Ph',
},
}
self.assertEqual(expected, pmod('Ph'))
self.assertEqual(expected, result.asDict())
def _help_test_non_standard_namespace(self, statement):
result = self.parser.parseString(statement)
expected = {
KIND: PMOD,
IDENTIFIER: entity('MOD', 'PhosRes'),
PMOD_CODE: 'Ser',
PMOD_POSITION: 473
}
self.assertEqual(expected, pmod(name='PhosRes', namespace='MOD', code='Ser', position=473))
self.assertEqual(expected, result.asDict())
def _help_test_pmod_with_residue(self, statement):
result = self.parser.parseString(statement)
expected = {
KIND: PMOD,
IDENTIFIER: {
NAMESPACE: BEL_DEFAULT_NAMESPACE,
NAME: 'Ph',
},
PMOD_CODE: 'Ser',
}
self.assertEqual(expected, pmod('Ph', code='Ser'))
self.assertEqual(expected, result.asDict())
def test_transphosphorylation():
egfr = Agent('EGFR', db_refs={'HGNC': id('EGFR')})
egfr_dimer = Agent('EGFR', bound_conditions=[BoundCondition(egfr)],
db_refs={'HGNC': id('EGFR')})
stmt = Transphosphorylation(egfr_dimer, 'Y', '1173')
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert belgraph.number_of_nodes() == 3
assert belgraph.number_of_edges() == 3
egfr_dimer_node = complex_abundance([egfr_dsl, egfr_dsl])
egfr_phos_node = egfr_dsl.with_variants(pmod('Ph', 'Tyr', 1173))
edge_data = get_edge_data(belgraph, egfr_dimer_node, egfr_phos_node)
assert edge_data == {pc.RELATION: pc.DIRECTLY_INCREASES}
def _help_test_pmod_full(self, statement):
result = self.parser.parseString(statement)
expected = {
KIND: PMOD,
IDENTIFIER: {
NAMESPACE: BEL_DEFAULT_NAMESPACE,
NAME: 'Ph',
},
PMOD_CODE: 'Ser',
PMOD_POSITION: 473
}
self.assertEqual(expected, pmod('Ph', code='Ser', position=473))
self.assertEqual(expected, result.asDict())
def test_update_matrix_inhibition_phosphorylation(self):
"""Test updating the matrix with an inhibition phosphorylation."""
sub = protein(namespace='HGNC', name='A', identifier='1')
obj = protein(namespace='HGNC', name='B', identifier='2', variants=[pmod('Ph')])
index = {'A', 'B'}
test_dict = {}
test_matrix = DataFrame(0, index=index, columns=index)
test_dict["inhibition_phosphorylation"] = test_matrix
update_spia_matrices(test_dict, sub, obj, {'relation': 'decreases'})
self.assertEqual(test_dict["inhibition_phosphorylation"]['A']['B'], 1)
self.assertEqual(test_dict["inhibition_phosphorylation"]['A']['A'], 0)
self.assertEqual(test_dict["inhibition_phosphorylation"]['B']['A'], 0)
self.assertEqual(test_dict["inhibition_phosphorylation"]['B']['B'], 0)
def test_complex_with_pmod():
sos1_phos = Agent('SOS1',
mods=[ModCondition('phosphorylation', 'Y', '100')],
db_refs={'HGNC': id('SOS1')})
grb2 = Agent('GRB2', db_refs={'HGNC': id('GRB2')})
egfr = Agent('EGFR', db_refs={'HGNC': id('EGFR')})
stmt = Complex([sos1_phos, grb2, egfr])
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert belgraph.number_of_nodes() == 5
assert belgraph.number_of_edges() == 4
egfr_grb2_sos_phos_tyr_100 = complex_abundance(
[egfr_dsl, grb2_dsl,
sos1_dsl.with_variants(pmod('Ph', 'Tyr', 100))])
assert sos1_dsl in belgraph
assert egfr_grb2_sos_phos_tyr_100 in belgraph
for member in egfr_grb2_sos_phos_tyr_100.members:
assert member in belgraph
def test_complex_with_pmod():
sos1_phos = Agent('SOS1',
mods=[ModCondition('phosphorylation', 'Y', '100')],
db_refs={'HGNC': id('SOS1')})
grb2 = Agent('GRB2', db_refs={'HGNC': id('GRB2')})
egfr = Agent('EGFR', db_refs={'HGNC': id('EGFR')})
stmt = Complex([sos1_phos, grb2, egfr])
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert belgraph.number_of_nodes() == 5
assert belgraph.number_of_edges() == 4
egfr_grb2_sos_phos_tyr_100 = complex_abundance([
egfr_dsl,
grb2_dsl,
sos1_dsl.with_variants(pmod('Ph', 'Tyr', 100))
])
assert sos1_dsl in belgraph
assert egfr_grb2_sos_phos_tyr_100 in belgraph
for member in egfr_grb2_sos_phos_tyr_100.members:
assert member in belgraph
def test_negativeCorrelation_withObjectVariant(self):
"""
3.2.1 http://openbel.org/language/web/version_2.0/bel_specification_version_2.0.html#XnegCor
Test phosphoralation tag"""
statement = 'kin(p(SFAM:"GSK3 Family")) neg p(HGNC:MAPT,pmod(P))'
result = self.parser.relation.parseString(statement)
expected_dict = {
SUBJECT: {
MODIFIER: ACTIVITY,
EFFECT: {
NAME: 'kin',
NAMESPACE: BEL_DEFAULT_NAMESPACE
},
TARGET: {
FUNCTION: PROTEIN,
NAMESPACE: 'SFAM',
NAME: 'GSK3 Family'
}
},
RELATION: NEGATIVE_CORRELATION,
OBJECT: {
FUNCTION: PROTEIN,
NAMESPACE: 'HGNC',
NAME: 'MAPT',
VARIANTS: [pmod('Ph')]
}
}
self.assertEqual(expected_dict, result.asDict())
sub = PROTEIN, 'SFAM', 'GSK3 Family'
self.assertHasNode(sub)
obj = PROTEIN, 'HGNC', 'MAPT', (PMOD, (BEL_DEFAULT_NAMESPACE, 'Ph'))
self.assertHasNode(obj)
self.assertHasEdge(sub, obj, relation=expected_dict[RELATION])
self.assertHasEdge(obj, sub, relation=expected_dict[RELATION])
def test_gef():
gef = Agent('SOS1', mods=[ModCondition('phosphorylation')],
db_refs={'HGNC': '11187'})
ras = Agent('KRAS', db_refs={'HGNC': '6407'})
stmt = Gef(gef, ras)
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert len(belgraph) == 3
assert belgraph.number_of_edges() == 2
gef_reference_node = protein(namespace='HGNC', name='SOS1')
gef_node = gef_reference_node.with_variants(pmod('Ph'))
assert gef_reference_node in belgraph
assert gef_node in belgraph
assert kras_node in belgraph
edge_data = get_edge_data(belgraph, gef_node, kras_node)
edge = {
pc.RELATION: pc.DIRECTLY_INCREASES,
pc.SUBJECT: activity('gef'),
pc.OBJECT: activity('gtp')
}
assert edge_data == edge
def test_gap():
gap = Agent('RASA1', mods=[ModCondition('phosphorylation')],
db_refs={'HGNC': '9871'})
ras = Agent('KRAS', db_refs={'HGNC': '6407'})
stmt = Gap(gap, ras)
pba = pa.PybelAssembler([stmt])
belgraph = pba.make_model()
assert len(belgraph) == 3
assert belgraph.number_of_edges() == 2
gap_reference_node = protein(namespace='HGNC', name='RASA1')
gap_node = gap_reference_node.with_variants(pmod('Ph'))
ras_node = protein(namespace='HGNC', name='KRAS')
assert gap_reference_node in belgraph
assert gap_node in belgraph
assert ras_node in belgraph
edge_data = get_edge_data(belgraph, gap_node, ras_node)
edge = {
pc.RELATION: pc.DIRECTLY_DECREASES,
pc.SUBJECT: activity('gap'),
pc.OBJECT: activity('gtp')
}
assert edge_data == edge
def test_protein_pmod(self):
node = protein(name='PLCG1',
namespace='HGNC',
variants=[pmod(name='Ph', code='Tyr')])
self.assertEqual('p(HGNC:PLCG1, pmod(Ph, Tyr))', str(node))
from pybel import BELGraph
from pybel.constants import CITATION_REFERENCE, CITATION_TYPE, CITATION_TYPE_OTHER, HAS_VARIANT
from pybel.dsl import (
abundance, activity, bioprocess, complex_abundance, gene, gmod, named_complex_abundance, pathology, pmod, protein,
protein_fusion, protein_substitution, reaction, rna,
)
example_graph = BELGraph()
example_graph.namespace_url['HGNC'] = ''
example_graph.namespace_pattern['DBSNP'] = r'^rs\d+$'
example_graph.annotation_url['Species'] = ''
example_graph.annotation_pattern['Number'] = r'^\d+$'
example_graph.annotation_list['Confidence'] = {'High', 'Low'}
ptk2 = protein(namespace='HGNC', name='PTK2', variants=[pmod('Ph', 'Tyr', 925)])
mapk1 = protein(namespace='HGNC', name='MAPK1')
mapk3 = protein(namespace='HGNC', name='MAPK3')
grb2 = protein(namespace='HGNC', name='GRB2')
sos1 = protein(namespace='HGNC', name='SOS1')
ptk2_rgb2_sos1 = complex_abundance([mapk1, grb2, sos1])
ras_family = protein(namespace='SFAM', name='RAS Family')
pi3k_complex = named_complex_abundance(namespace='SFAM', name='p85/p110 PI3Kinase Complex')
kinase_activity = activity('kin')
catalytic_activity = activity('cat')
gtp_activity = activity('gtp')
c1 = '10446041'
e1 = "FAK also combines with, and may activate, phosphoinositide 3-OH kinase (PI 3-kinase), either directly or " \
from pybel.testing.utils import n
from pybel_tools.assembler.reified_graph.assembler import (
ACTIVATES,
DEGRADATES,
FRAGMENTS,
INCREASES_ABUNDANCE,
PHOSPHORYLATES,
PROMOTES_TRANSLATION,
REIF_OBJECT,
REIF_SUBJECT,
reify_bel_graph,
)
cdk5 = protein('HGNC', 'CDK5', '1774')
gsk3b = protein('HGNC', 'GSK3B', '4617')
p_tau = protein('HGNC', 'MAPT', '6893', variants=pmod('Ph'))
# act(p(HGNC:FAS), ma(cat)) increases act(p(HGNC:CASP8), ma(cat))
fas = protein('HGNC', 'FAS', '11920')
casp8 = protein('HGNC', 'CASP8', '1509')
# a(CHEBI:oxaliplatin) increases a(MESHC:"Reactive Oxygen Species")
oxaliplatin = abundance('CHEBI', 'oxaliplatin', '31941')
reactive_o_species = abundance('MESH', 'Reactive Oxygen Species', 'D017382')
# p(HGNC:MYC) decreases r(HGNC:CCNB1)
@unittest.skip('Seems like tests were not working well at first')
class TestAssembleReifiedGraph(unittest.TestCase):
"""Test assembly of reified graphs."""
bioprocess,
gene,
gmod,
named_complex_abundance,
pathology,
pmod,
protein_fusion,
protein_substitution,
reaction,
rna,
)
from pybel.utils import citation_dict
example_graph = BELGraph()
ptk2 = Protein(namespace='hgnc', name='PTK2', variants=pmod('Ph', 'Tyr', 925))
mapk1 = Protein(namespace='hgnc', name='MAPK1')
mapk3 = Protein(namespace='hgnc', name='MAPK3')
grb2 = Protein(namespace='hgnc', name='GRB2')
sos1 = Protein(namespace='hgnc', name='SOS1')
ptk2_rgb2_sos1 = ComplexAbundance([mapk1, grb2, sos1])
ras_family = Protein(namespace='fplx', name='RAS')
pi3k_complex = named_complex_abundance(namespace='fplx',
name='p85/p110 PI3Kinase Complex')
kinase_activity = activity('kin')
catalytic_activity = activity('cat')
gtp_activity = activity('gtp')
c1 = '10446041'
import unittest
from pybel import BELGraph
from pybel.constants import DIRECTLY_INCREASES
from pybel.dsl import gene, mirna, pathology, pmod, protein, rna
from pybel.struct.mutation.collapse import collapse_all_variants, collapse_nodes, collapse_to_genes
from pybel.testing.utils import n
HGNC = 'HGNC'
GO = 'GO'
CHEBI = 'CHEBI'
g1 = gene(HGNC, '1')
r1 = rna(HGNC, '1')
p1 = protein(HGNC, '1')
p1_phosphorylated = protein(HGNC, '1', variants=[pmod('Ph')])
g2 = gene(HGNC, '2')
r2 = rna(HGNC, '2')
p2 = protein(HGNC, '2')
g3 = gene(HGNC, '3')
r3 = rna(HGNC, '3')
p3 = protein(HGNC, '3')
g4 = gene(HGNC, '4')
m4 = mirna(HGNC, '4')
p5 = pathology(GO, '5')
is_degraded,
is_gene,
is_pathology,
is_protein,
is_translocated,
keep_node_permissive,
node_exclusion_predicate_builder,
node_inclusion_predicate_builder,
not_pathology,
)
from pybel.testing.utils import n
p1 = protein(name='BRAF', namespace='HGNC')
p2 = protein(name='BRAF',
namespace='HGNC',
variants=[hgvs('p.Val600Glu'), pmod('Ph')])
p3 = protein(name='APP',
namespace='HGNC',
variants=fragment(start=672, stop=713))
p4 = protein(name='2', namespace='HGNC')
g1 = gene(name='BRAF', namespace='HGNC', variants=gmod('Me'))
class TestNodePredicates(unittest.TestCase):
"""Tests for node predicates."""
def test_none_data(self):
"""Test permissive node predicate with a node data dictionary."""
self.assertTrue(keep_node_permissive(p1))
def test_none(self):
from pybel.struct.filters import false_node_predicate, true_node_predicate
from pybel.struct.filters.edge_predicate_builders import build_relation_predicate
from pybel.struct.filters.edge_predicates import (
edge_has_activity, edge_has_annotation, edge_has_degradation,
edge_has_translocation, has_authors, has_polarity, has_provenance, has_pubmed, is_associative_relation,
is_causal_relation, is_direct_causal_relation,
)
from pybel.struct.filters.node_predicates import (
has_activity, has_causal_in_edges, has_causal_out_edges, has_fragment, has_gene_modification, has_hgvs,
has_protein_modification, has_variant, is_abundance, is_causal_central, is_causal_sink, is_causal_source,
is_degraded, is_gene, is_pathology, is_protein, is_translocated, none_of, not_pathology, one_of,
)
from pybel.testing.utils import n
p1 = protein(name='BRAF', namespace='HGNC')
p2 = protein(name='BRAF', namespace='HGNC', variants=[hgvs('p.Val600Glu'), pmod('Ph')])
p3 = protein(name='APP', namespace='HGNC', variants=fragment(start=672, stop=713))
p4 = protein(name='2', namespace='HGNC')
g1 = gene(name='BRAF', namespace='HGNC', variants=gmod('Me'))
class TestNodePredicates(unittest.TestCase):
"""Tests for node predicates."""
def test_none_data(self):
"""Test permissive node predicate with a node data dictionary."""
self.assertTrue(true_node_predicate(p1))
self.assertFalse(false_node_predicate(p1))
def test_none(self):
def add_simple_edge(graph, u, v, relation_type):
"""Add corresponding edge type to BEL graph.
:param pybel.BELGraph graph: BEL Graph
:param u: source node
:param v: target node
:param list relation_type: source ID, target ID and relation types
"""
# Check if multiple relation subtypes present
if isinstance(relation_type, list):
# Check if protein modification is a relation subtype
if relation_type[1] in {
'phosphorylation', 'glycosylation', 'ubiquitination',
'methylation'
}:
# If the object is a gene, miRNA, RNA, or protein, add protein modification
if isinstance(v, CentralDogma):
v = v.with_variants(pmod(KEGG_MODIFICATIONS[relation_type[1]]))
# Add increases edge if pmod subtype is coupled with activation subtype
if relation_type[0] == 'activation':
graph.add_increases(
u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
subject_modifier=activity() if isinstance(
u, ACTIVITY_ALLOWED_MODIFIERS) else None,
# Add the activity function if subject is one of the following nodes (BEL 2.0 specifications)
annotations={},
)
return
# Add decreases edge if pmod subtype is coupled with inhibition subtype
elif relation_type[0] == 'inhibition':
graph.add_decreases(
u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
subject_modifier=activity() if isinstance(
u, ACTIVITY_ALLOWED_MODIFIERS) else None,
# Add the activity function if subject is one of the following nodes (BEL 2.0 specifications)
annotations={},
)
return
# Found multiple relationship which cannot be combined into one logic
# (e.g., ['inhibition', 'indirect effect'])
else:
# Create all relationships in the list
for relation in relation_type:
add_simple_edge(graph, u, v, relation)
return
# Found multiple relationship which cannot be combined into one logic (e.g., ['inhibition', 'indirect effect'])
else:
# Create all relationships in the list
for relation in relation_type:
add_simple_edge(graph, u, v, relation)
return
"""Handle differently the relationships"""
# If only one pmod relation subtype
if relation_type in {
'phosphorylation', 'glycosylation', 'ubiquitination', 'methylation'
}:
# If the object is a gene, miRNA, RNA, or protein, add protein modification
if isinstance(v, CentralDogma):
v = v.with_variants(pmod(KEGG_MODIFICATIONS[relation_type]))
graph.add_increases(
u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
subject_modifier=activity() if isinstance(
u, ACTIVITY_ALLOWED_MODIFIERS) else None,
annotations={},
)
return
# Subject activity decreases protein modification (i.e. dephosphorylation) of object
elif relation_type == 'dephosphorylation':
# If the object is a gene, miRNA, RNA, or protein, add protein modification
if isinstance(v, CentralDogma):
v = v.with_variants(pmod('Ph'))
graph.add_decreases(
u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
subject_modifier=activity() if isinstance(
u, ACTIVITY_ALLOWED_MODIFIERS) else None,
annotations={},
)
return
# Subject increases activity of object
elif relation_type == 'activation':
graph.add_increases(
u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
object_modifier=activity() if isinstance(
v, ACTIVITY_ALLOWED_MODIFIERS) else None,
annotations={},
)
return
# Catalytic activity of subject increases transformation of reactant(s) to product(s)
elif relation_type in {'reversible', 'irreversible'}:
graph.add_increases(
u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
subject_modifier=activity('cat') if isinstance(
u, ACTIVITY_ALLOWED_MODIFIERS) else None,
annotations={},
)
return
# Subject decreases activity of object
elif relation_type == 'inhibition':
graph.add_decreases(
u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
object_modifier=activity() if isinstance(
v, ACTIVITY_ALLOWED_MODIFIERS) else None,
annotations={},
)
return
# Indirect effect and binding/association are noted to be equivalent relation types
elif relation_type in {
'indirect effect', 'binding/association', 'compound'
}:
graph.add_association(u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
annotations={})
return
# Subject increases expression of object
elif relation_type == 'expression':
# Expression object is converted to RNA abundance
if isinstance(v, CentralDogma):
v = v.get_rna()
graph.add_increases(u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
annotations={})
return
# Subject decreases expression of object
elif relation_type == 'repression':
# Repression object is converted to RNA abundance
if isinstance(v, CentralDogma):
v = v.get_rna()
graph.add_decreases(u,
v,
citation=KEGG_CITATION,
evidence='Extracted from KEGG',
annotations={})
return
elif relation_type in {
'dissociation', 'hidden compound', 'missing interaction',
'state change'
}:
return
raise ValueError(
f'Unexpected relation type {relation_type} between {u} and {v}')
import networkx as nx
import pybel.constants as pc
from pybel.dsl import abundance, activity, bioprocess, \
complex_abundance, hgvs, pmod, protein, reaction
from indra.assemblers.pybel import assembler as pa
from indra.databases import hgnc_client
from indra.statements import *
def id(gene_name):
return hgnc_client.get_hgnc_id(gene_name)
phos_dsl = pmod('Ph', 'Ser', 218)
ub_dsl = pmod('Ub', 'Ser', 218)
egfr_phos_dsl = pmod('Ph', 'Tyr', 1173)
braf_dsl = protein(namespace='HGNC', name='BRAF', identifier='1097')
map2k1_dsl = protein(namespace='HGNC', name='MAP2K1', identifier='6840')
tp53_dsl = protein(namespace='HGNC', name='TP53', identifier='11998')
mdm2_dsl = protein(namespace='HGNC', name='MDM2', identifier='6973')
egfr_dsl = protein(namespace='HGNC', name='EGFR', identifier='3236')
chebi_17534 = abundance(namespace='CHEBI', name='D-glucose',
identifier='17634')
chebi_4170 = abundance(namespace='CHEBI', name='D-glucopyranose 6-phosphate',
identifier='4170')
chebi_17534_to_4170 = reaction(chebi_17534, chebi_4170)
Protein('HGNC', 'CXCR3'),
Abundance('CHEBI', 'acrolein'),
Protein('HGNC', 'IDO2'),
Pathology('MESHD', 'Pulmonary Disease, Chronic Obstructive'),
Protein('HGNC', 'IFNG'),
Protein('HGNC', 'TNFRSF4'),
Protein('HGNC', 'CTLA4'),
Protein('HGNC', 'GZMA'),
Protein('HGNC', 'PRF1'),
Protein('HGNC', 'TNF'),
Protein('SFAM', 'Chemokine Receptor Family'),
ComplexAbundance([Protein('HGNC', 'CD8A'),
Protein('HGNC', 'CD8B')]),
ComplexAbundance([Protein('HGNC', 'CD8A'),
Protein('HGNC', 'CD8B')]),
Protein('HGNC', 'PLCG1', variants=pmod('Ph', 'Tyr')),
Protein('EGID', '21577'),
}
jgif_expected_edges = [
(calcium, calcineurin_complex, {
RELATION: DIRECTLY_INCREASES,
EVIDENCE:
'NMDA-mediated influx of calcium led to activated of the calcium-dependent phosphatase calcineurin and the subsequent dephosphorylation and activation of the protein-tyrosine phosphatase STEP',
CITATION: {
CITATION_TYPE: CITATION_TYPE_PUBMED,
CITATION_REFERENCE: '12483215'
},
OBJECT: {
MODIFIER: ACTIVITY,
EFFECT: {
import networkx as nx
import pybel.constants as pc
from indra.assemblers.pybel import assembler as pa
from indra.databases import hgnc_client
from indra.statements import *
from pybel.dsl import abundance, activity, bioprocess, complex_abundance, hgvs, pmod, protein, reaction
def id(gene_name):
return hgnc_client.get_hgnc_id(gene_name)
phos_dsl = pmod('Ph', 'Ser', 218)
ub_dsl = pmod('Ub', 'Ser', 218)
egfr_phos_dsl = pmod('Ph', 'Tyr', 1173)
braf_dsl = protein(namespace='HGNC', name='BRAF')
map2k1_dsl = protein(namespace='HGNC', name='MAP2K1')
tp53_dsl = protein(namespace='HGNC', name='TP53')
mdm2_dsl = protein(namespace='HGNC', name='MDM2')
egfr_dsl = protein(namespace='HGNC', name='EGFR')
chebi_17534 = abundance(namespace='CHEBI', name='17634')
chebi_4170 = abundance(namespace='CHEBI', name='4170')
chebi_17534_to_4170 = reaction(chebi_17534, chebi_4170)
grb2_dsl = protein(namespace='HGNC', name='GRB2')
sos1_dsl = protein(namespace='HGNC', name='SOS1')
sos1_phosphorylated_dsl = sos1_dsl.with_variants(pmod('Ph'))
kras_node = protein(namespace='HGNC', name='KRAS')
