def test_types():
G = NdexGraph()
n = G.add_new_node('Node with Types')
n1 = G.add_new_node('A')
n2 = G.add_new_node('B')
G.add_edge_between(n, n1)
G.add_edge_between(n, n2)
G.set_name('Test Types')
G.node[n]['string'] = 'mystring'
G.node[n]['bool'] = True
G.node[n]['int'] = 5
G.node[n]['double'] = 2.5
# Python3 doesn't support Long (you cannot have a = 5L);
# If we need an integer being a long in Python 2 and still be compatible with Python 3,
# we need to define up a long variable to be the same as the int class under Python 3,
# and it can then be used explicitly to make sure the integer is a long.
#
# it is taken from http: //python3porting.com/noconv.html
if sys.version_info.major == 3:
long = int
# long(5) will be 5L in Python 2 and just int 5 (which is long) in Python 3
G.node[n]['long'] = long(5)
G.node[n]['string_list'] = ['mystring', 'myotherstring']
G.node[n]['bool_list'] = [False, True]
G.node[n]['int_list'] = [5, -20]
G.node[n]['double_list'] = [2.5, 3.7]
# long(5), long(75) will be 5L, 75L in Python 2 and just int 5, 75 (which is long) in Python 3
G.node[n]['long_list'] = [long(5), long(75)]
G.write_to('temp_test_type.cx')
def scratch_test():
G = NdexGraph()
G.add_new_node('1')
G.add_new_node('2')
G.add_edge_between(1,2)
G.set_name('scratch-test1 - jing')
print(G.graph())
def scratch_test():
G = NdexGraph()
G.add_new_node('1')
G.add_new_node('2')
G.add_edge_between(1, 2)
G.set_name('scratch-test1 - jing')
print(G.graph())
def test_layout():
nx_G = nx.complete_graph(11)
# nx_G.remove_node(0)
# nx_G.remove_node(1)
G = NdexGraph()
G.add_new_node('1')
G.add_new_node('2')
G.add_new_node('3')
G.add_new_node('4')
G.add_new_node('5')
G.add_new_node('6')
G.add_edge_between(1, 3)
G.add_edge_between(1, 4)
G.add_edge_between(2, 3)
G.add_edge_between(2, 4)
G.add_edge_between(3, 5)
G.add_edge_between(3, 6)
G.add_edge_between(4, 5)
G.add_edge_between(
4, 6, interaction='AAABBBCCC') #attr_dict={'interaction':'testing'})
initial_pos = {
1: (0.0, 1.0),
2: (0.0, 0.0),
# 3: (0.5, 0.5),
# 4: (0.5, 0.5),
5: (1.0, 1.0),
6: (1.0, 0.0),
}
fixed = [1, 2, 5, 6]
# G.add_new_node('3')
# G.add_new_node('4')
# G.add_new_node('5')
G.pos = nx.spring_layout(G.to_undirected(), pos=initial_pos, fixed=fixed)
print(G.pos)
G.set_name('spring_layout undirected')
def test_layout():
nx_G = nx.complete_graph(11)
# nx_G.remove_node(0)
# nx_G.remove_node(1)
G = NdexGraph()
G.add_new_node('1')
G.add_new_node('2')
G.add_new_node('3')
G.add_new_node('4')
G.add_new_node('5')
G.add_new_node('6')
G.add_edge_between(1, 3)
G.add_edge_between(1, 4)
G.add_edge_between(2, 3)
G.add_edge_between(2, 4)
G.add_edge_between(3, 5)
G.add_edge_between(3, 6)
G.add_edge_between(4, 5)
G.add_edge_between(4, 6, interaction='AAABBBCCC') #attr_dict={'interaction':'testing'})
initial_pos = {
1: (0.0, 1.0),
2: (0.0, 0.0),
# 3: (0.5, 0.5),
# 4: (0.5, 0.5),
5: (1.0, 1.0),
6: (1.0, 0.0),
}
fixed = [1,2,5,6]
# G.add_new_node('3')
# G.add_new_node('4')
# G.add_new_node('5')
G.pos = nx.spring_layout(G.to_undirected(), pos=initial_pos, fixed=fixed)
print(G.pos)
G.set_name('spring_layout undirected')
def test_types():
G = NdexGraph()
n = G.add_new_node('Node with Types')
n1 = G.add_new_node('A')
n2 = G.add_new_node('B')
G.add_edge_between(n, n1)
G.add_edge_between(n, n2)
G.set_name('Test Types')
G.node[n]['string'] = 'mystring'
G.node[n]['bool'] = True
G.node[n]['int'] = 5
G.node[n]['double'] = 2.5
# Python3 doesn't support Long (you cannot have a = 5L);
# If we need an integer being a long in Python 2 and still be compatible with Python 3,
# we need to define up a long variable to be the same as the int class under Python 3,
# and it can then be used explicitly to make sure the integer is a long.
#
# it is taken from http: //python3porting.com/noconv.html
if sys.version_info.major == 3:
long = int
# long(5) will be 5L in Python 2 and just int 5 (which is long) in Python 3
G.node[n]['long'] = long(5)
G.node[n]['string_list'] = ['mystring','myotherstring']
G.node[n]['bool_list'] = [False, True]
G.node[n]['int_list'] = [5, -20]
G.node[n]['double_list'] = [2.5, 3.7]
# long(5), long(75) will be 5L, 75L in Python 2 and just int 5, 75 (which is long) in Python 3
G.node[n]['long_list'] = [long(5), long(75)]
G.write_to('temp_test_type.cx')
def test_types():
G = NdexGraph()
n = G.add_new_node('Node with Types')
n1 = G.add_new_node('A')
n2 = G.add_new_node('B')
G.add_edge_between(n, n1)
G.add_edge_between(n, n2)
G.set_name('Test Types')
G.node[n]['string'] = 'mystring'
G.node[n]['bool'] = True
G.node[n]['int'] = 5
G.node[n]['double'] = 2.5
G.node[n]['long'] = 5L
G.node[n]['string_list'] = ['mystring', 'myotherstring']
G.node[n]['bool_list'] = [False, True]
G.node[n]['int_list'] = [5, -20]
G.node[n]['double_list'] = [2.5, 3.7]
G.node[n]['long_list'] = [5L, 75L]
G.write_to('temp_test_type.cx')
G.upload_to('http://test.ndexbio.org', 'scratch', 'scratch')
def test_types():
G = NdexGraph()
n = G.add_new_node('Node with Types')
n1 = G.add_new_node('A')
n2 = G.add_new_node('B')
G.add_edge_between(n, n1)
G.add_edge_between(n, n2)
G.set_name('Test Types')
G.node[n]['string'] = 'mystring'
G.node[n]['bool'] = True
G.node[n]['int'] = 5
G.node[n]['double'] = 2.5
G.node[n]['long'] = 5L
G.node[n]['string_list'] = ['mystring','myotherstring']
G.node[n]['bool_list'] = [False, True]
G.node[n]['int_list'] = [5, -20]
G.node[n]['double_list'] = [2.5, 3.7]
G.node[n]['long_list'] = [5L, 75L]
G.write_to('temp_test_type.cx')
G.upload_to('http://test.ndexbio.org', 'scratch', 'scratch')
def test_complex_layout_create():
G = NdexGraph()
s = G.add_new_node('Source', type='Source')
t = G.add_new_node('Target', type='Target')
f = G.add_new_node('Forward', type='Forward')
r = G.add_new_node('Reverse', type='Reverse')
b = G.add_new_node('Both', type='Both')
G.add_edge_between(s, f)
G.add_edge_between(f, t)
G.add_edge_between(t, r)
G.add_edge_between(r, s)
G.add_edge_between(s, b)
G.add_edge_between(b, t)
G.add_edge_between(t, b)
G.add_edge_between(b, s)
G.set_name('layout template')
G.upload_to('http://public.ndexbio.org', 'scratch', 'scratch')
def test_complex_layout_create():
G = NdexGraph()
s = G.add_new_node('Source', type='Source')
t = G.add_new_node('Target', type='Target')
f = G.add_new_node('Forward', type='Forward')
r = G.add_new_node('Reverse', type='Reverse')
b = G.add_new_node('Both', type='Both')
G.add_edge_between(s, f)
G.add_edge_between(f, t)
G.add_edge_between(t, r)
G.add_edge_between(r, s)
G.add_edge_between(s, b)
G.add_edge_between(b, t)
G.add_edge_between(t, b)
G.add_edge_between(b, s)
G.set_name('layout template')
G.upload_to('http://public.ndexbio.org', 'scratch', 'scratch')
def test_template():
template_id = 'f35fbfd3-4918-11e6-a5c7-06603eb7f303'
G = NdexGraph()
G.add_new_node('1', color='green')
G.add_new_node('2', color='red')
G.add_new_node('3', color='green')
G.add_new_node('4', color='green')
G.add_new_node('5', color='green')
G.add_new_node('6', color='red')
G.add_edge_between(1, 2)
G.add_edge_between(2, 3)
G.add_edge_between(3, 4)
G.add_edge_between(4, 5)
G.add_edge_between(5, 6)
toolbox.apply_template(G, template_id)
G.set_name('template apply')
G.upload_to('http://public.ndexbio.org', 'scratch', 'scratch')
def test_template():
template_id = 'f35fbfd3-4918-11e6-a5c7-06603eb7f303'
G = NdexGraph()
G.add_new_node('1', color='green')
G.add_new_node('2', color='red')
G.add_new_node('3', color='green')
G.add_new_node('4', color='green')
G.add_new_node('5', color='green')
G.add_new_node('6', color='red')
G.add_edge_between(1, 2)
G.add_edge_between(2, 3)
G.add_edge_between(3, 4)
G.add_edge_between(4, 5)
G.add_edge_between(5, 6)
toolbox.apply_template(G, template_id)
G.set_name('template apply')
G.upload_to('http://public.ndexbio.org', 'scratch', 'scratch')
for x in name:
if x.isdigit():
i = i + 1
G.add_new_node(
name, {"Number of Integers": sum(nd.isdigit() for nd in name)})
#adding edges
counter = 0
j = 1
iteration = 0
while j < 10:
counter = j + 1
while counter < 11:
iteration = iteration + 1
G.add_edge_between(j, counter)
similar = 0
for jChar in G.get_node_name_by_id(j):
for counterChar in G.get_node_name_by_id(counter):
if jChar == counterChar:
similar = similar + 1
counter = counter + 1
G.set_edge_attribute(iteration, "Similar Characters", similar)
j = j + 1
#print statements
#edgesx = combinations(list, 2)
print "There are " + str(iteration) + " edges"
dictEdges = {}
def create_similarity_map(name, e_set, min_subsumption, id_attribute="genes", max_edges=5):
similarity_graph = NdexGraph()
similarity_graph.set_name(name)
set_name_to_node_id_map = {}
id_sets = {}
for network_id in e_set.id_set_map:
id_set_object = e_set.id_set_map[network_id]
network_name = id_set_object.name
id_set = id_set_object.set
id_sets[network_id] = id_set
att = {id_attribute: list(id_set)}
node_id = similarity_graph.add_new_node(network_name, att)
set_name_to_node_id_map[network_id] = node_id
source_similarities = {}
for network_id_1 in id_sets.keys():
source_node_id = set_name_to_node_id_map[network_id_1]
list_1 = list(id_sets[network_id_1])
set_1_size = len(list_1)
similarities = []
for network_id_2 in id_sets.keys():
if network_id_1 != network_id_2:
set_1 = id_sets[network_id_1]
set_2 = id_sets[network_id_2]
target_node_id = set_name_to_node_id_map[network_id_2]
list_2 = list(id_sets[network_id_2])
set_2_size = len(list_2)
overlap = list(set_1.intersection(set_2))
size_overlap=len(overlap)
if size_overlap != 0:
subsumes = size_overlap/set_2_size
#subsumes_1 = size_overlap/set_2_size
#subsumes_2 = size_overlap/set_1_size
#subsumes = min(subsumes_1, subsumes_2)
if size_overlap > 3:
print "overlap: %s %s" % (size_overlap, overlap)
similarity = {"source_node_id": source_node_id,
"target_node_id": target_node_id,
"subsumes": subsumes}
similarity["atts"] = {"subsumes": subsumes,
"overlap": overlap,
"overlap_size": size_overlap}
similarities.append(similarity)
else:
print "no overlap"
# rank the similarities
similarities = sorted(similarities, key=operator.itemgetter('subsumes'), reverse=True)
source_similarities[network_id_1] = similarities
# always include the most similar node to make sure that each node has at least one edge and the graph is connected
# don't connect more than max_edges
for network_id, similarities in source_similarities.iteritems():
count = 0
for similarity in similarities:
if count >= max_edges:
break
if count == 0 or similarity["subsumes"] > min_subsumption:
atts = similarity["atts"]
source_node_id = similarity["source_node_id"]
target_node_id = similarity["target_node_id"]
similarity_graph.add_edge_between(source_node_id, target_node_id, attr_dict=atts)
count = count + 1
return similarity_graph
def test_complex_layout():
G = NdexGraph()
s1 = G.add_new_node('S1', st_layout='Source')
s2 = G.add_new_node('S2', st_layout='Source')
s3 = G.add_new_node('S3', st_layout='Source')
t1 = G.add_new_node('T1', st_layout='Target')
t2 = G.add_new_node('T2', st_layout='Target')
f1 = G.add_new_node('F1', st_layout='Forward')
f2 = G.add_new_node('F2', st_layout='Forward')
f3 = G.add_new_node('F3', st_layout='Forward')
f4 = G.add_new_node('F4', st_layout='Forward')
f5 = G.add_new_node('F5', st_layout='Forward')
r1 = G.add_new_node('R1', st_layout='Reverse')
r2 = G.add_new_node('R2', st_layout='Reverse')
r3 = G.add_new_node('R3', st_layout='Reverse')
b = G.add_new_node('B1', st_layout='Both')
G.add_edge_between(r3,r2)
G.add_edge_between(r2,r1)
G.add_edge_between(r1,s1)
G.add_edge_between(r3,b)
G.add_edge_between(s2, f3)
G.add_edge_between(s3, f1)
G.add_edge_between(f1, f2)
G.add_edge_between(f2, t1)
G.add_edge_between(f1, b)
G.add_edge_between(f3, f4)
G.add_edge_between(f4, f5)
G.add_edge_between(f5, t2)
G.add_edge_between(b, t1)
G.add_edge_between(b, s2)
G.add_edge_between(b, r1)
G.add_edge_between(t1,r3)
toolbox.apply_source_target_layout(G)
template_id = 'd1856d17-4937-11e6-a5c7-06603eb7f303'
toolbox.apply_template(G, template_id)
G.set_name('experiment1')
G.upload_to('http://public.ndexbio.org', 'scratch', 'scratch')
def test_complex_layout():
G = NdexGraph()
s1 = G.add_new_node('S1', st_layout='Source')
s2 = G.add_new_node('S2', st_layout='Source')
s3 = G.add_new_node('S3', st_layout='Source')
t1 = G.add_new_node('T1', st_layout='Target')
t2 = G.add_new_node('T2', st_layout='Target')
f1 = G.add_new_node('F1', st_layout='Forward')
f2 = G.add_new_node('F2', st_layout='Forward')
f3 = G.add_new_node('F3', st_layout='Forward')
f4 = G.add_new_node('F4', st_layout='Forward')
f5 = G.add_new_node('F5', st_layout='Forward')
r1 = G.add_new_node('R1', st_layout='Reverse')
r2 = G.add_new_node('R2', st_layout='Reverse')
r3 = G.add_new_node('R3', st_layout='Reverse')
b = G.add_new_node('B1', st_layout='Both')
G.add_edge_between(r3, r2)
G.add_edge_between(r2, r1)
G.add_edge_between(r1, s1)
G.add_edge_between(r3, b)
G.add_edge_between(s2, f3)
G.add_edge_between(s3, f1)
G.add_edge_between(f1, f2)
G.add_edge_between(f2, t1)
G.add_edge_between(f1, b)
G.add_edge_between(f3, f4)
G.add_edge_between(f4, f5)
G.add_edge_between(f5, t2)
G.add_edge_between(b, t1)
G.add_edge_between(b, s2)
G.add_edge_between(b, r1)
G.add_edge_between(t1, r3)
toolbox.apply_source_target_layout(G)
template_id = 'd1856d17-4937-11e6-a5c7-06603eb7f303'
toolbox.apply_template(G, template_id)
G.set_name('experiment1')
G.upload_to('http://public.ndexbio.org', 'scratch', 'scratch')
def ebs_to_network(ebs, name="not named"):
G = NdexGraph()
G.set_name(name)
node_id_map = {}
identifier_citation_id_map = {}
node_mapping = {}
mg = mygene.MyGeneInfo()
# Create Nodes
# PARTICIPANT PARTICIPANT_TYPE PARTICIPANT_NAME UNIFICATION_XREF RELATIONSHIP_XREF
id_list = []
for node in ebs.get("node_table"):
if "PARTICIPANT" in node:
participant = node["PARTICIPANT"]
participant = clean_brackets(participant)
if participant is not None and '_HUMAN' in participant and node_mapping.get(
participant) is None:
id_list.append(participant)
url = 'https://biodbnet-abcc.ncifcrf.gov/webServices/rest.php/biodbnetRestApi.json?method=db2db&input=uniprot entry name&inputValues=' + ','.join(
id_list) + '&outputs=genesymbol&taxonId=9606&format=row'
look_up_req = requests.get(url)
look_up_json = look_up_req.json()
if look_up_json is not None:
for bio_db_item in look_up_json:
gene_symbol_mapping[bio_db_item.get(
'InputValue')] = bio_db_item.get('Gene Symbol')
node_mapping[bio_db_item.get('InputValue')] = bio_db_item.get(
'Gene Symbol')
for node in ebs.get("node_table"):
attributes = {}
if "PARTICIPANT" in node:
participant = node["PARTICIPANT"]
participant = clean_brackets(participant)
participant_symbol = ''
if participant is not None:
#participant = participant.replace('_HUMAN', '')
participant_symbol = gene_symbol_mapping.get(participant)
#=====================================
# if the node is not a protien type
# then skip mygene.info look up
#=====================================
if node.get('PARTICIPANT_TYPE') is not None and node.get(
'PARTICIPANT_TYPE') != 'ProteinReference':
participant_symbol = participant
if participant_symbol is None:
participant_symbol = mg.query(participant, species='human')
if participant_symbol is not None and participant_symbol.get(
'hits') is not None and len(
participant_symbol.get('hits')) > 0:
ph = participant_symbol.get('hits')[0]
if 'symbol' in ph:
participant_symbol = str(ph.get('symbol'))
gene_symbol_mapping[
participant] = participant_symbol
else:
participant_symbol = None
participant_name = None
if participant_symbol is not None:
node_name = participant_symbol
else:
node_name = participant
node_name_tmp = participant
if "PARTICIPANT_NAME" in node:
name = node["PARTICIPANT_NAME"]
if name is not None and len(name) > 0:
attributes["name"] = name
participant_name = name
node_type = _get_node_type(node.get("PARTICIPANT_TYPE"))
if "UNIFICATION_XREF" in node:
alias_string = node["UNIFICATION_XREF"]
if alias_string is not None and alias_string is not "":
aliases = alias_string.split(";")
attributes["alias"] = aliases
# attributes["represents"] = alias[0] - can't take first alias for ndexebs.
# Need to resolve uniprot primary id for the gene
if node_type == "Other":
node_type = layouts.aliases_to_node_type(aliases)
attributes["type"] = node_type
if node_name_tmp.startswith("CHEBI") and participant_name:
node_name = participant_name
# check mygene.info
#if there is a result use the gene symbol and add the uniprot id as an alias
node_id = G.add_new_node(node_name, attributes)
node_id_map[participant] = node_id
# Create Edges
# PARTICIPANT_A INTERACTION_TYPE PARTICIPANT_B INTERACTION_DATA_SOURCE
# INTERACTION_PUBMED_ID PATHWAY_NAMES MEDIATOR_IDS
for edge in ebs.get("edge_table"):
if "INTERACTION_TYPE" not in edge:
raise "No interaction type for edge " + str(edge)
if "PARTICIPANT_A" not in edge:
raise "No participant A in edge " + str(edge)
if "PARTICIPANT_B" not in edge:
raise "No participant B in edge " + str(edge)
attributes = {}
interaction = edge["INTERACTION_TYPE"]
attributes["interaction"] = interaction
if interaction in DIRECTED_INTERACTIONS:
attributes["directed"] = True
# handle pmids as edge attribute
if "INTERACTION_PUBMED_ID" in edge:
pmid_string = edge["INTERACTION_PUBMED_ID"]
if pmid_string is not None and pmid_string is not "":
pmids = pmid_string.split(";")
attributes["pmid"] = pmids
participant_a = clean_brackets(edge["PARTICIPANT_A"])
participant_b = clean_brackets(edge["PARTICIPANT_B"])
#if participant_a is not None:
# participant_a = participant_a.replace('_HUMAN', '')
#if participant_b is not None:
# participant_b = participant_b.replace('_HUMAN', '')
source_node_id = node_id_map.get(participant_a)
target_node_id = node_id_map.get(participant_b)
edge_id = G.add_edge_between(source_node_id,
target_node_id,
interaction=interaction,
attr_dict=attributes)
# handle pmids in CX citation and edgeCitation aspect representations in networkn network object
if "pmid" in attributes:
prefixed_pmids = []
for pmid in attributes["pmid"]:
prefixed_pmids.append("pmid:" + str(pmid))
for prefixed_pmid in prefixed_pmids:
if prefixed_pmid not in identifier_citation_id_map:
citation_id = G.add_citation(identifier=prefixed_pmid)
identifier_citation_id_map[prefixed_pmid] = citation_id
G.add_citation_to_edge(
edge_id, prefixed_pmid
) #add_edge_citation_ref(edge_id, citation_id)
else:
citation_id = identifier_citation_id_map[prefixed_pmid]
G.add_citation_to_edge(
edge_id, prefixed_pmid
) #add_edge_citation_ref(edge_id, citation_id)
# save gene_symbol_mapping to file (caching)
with open(join(current_directory, 'gene_symbol_mapping.json'),
'w') as outfile:
json.dump(gene_symbol_mapping, outfile, indent=4)
return G
