def toBel(dbId):
''' Convert to BEL formats '''
entity = getEntityData(dbId)
print('toBel dbId: ', dbId)
type = entity['schemaClass']
if (type == 'Compartment'):
bel = toBelCompartment(entity)
elif type == 'EntityCompartment':
bel = toBelEntityCompartment(entity)
elif type == 'OtherEntity':
bel = toBelOtherEntity(entity)
elif type == 'Polymer':
bel = toBelPolymer(entity)
elif type == 'Complex':
bel = toBelComplexComponents(entity) # alternative version - toBelComplexNamed()
elif type == 'SimpleEntity':
bel = toBelSimpleEntity(entity)
elif type == 'EntityWithAccessionedSequence':
bel = toBelEntityWithAccessionedSequence(entity)
elif 'Set' in type:
bel = toBelSets(entity)
elif 'GenomeEncodedEntity' in type:
bel = toBelGenomeEncodedEntity(entity)
elif 'CatalystActivity' in type:
bel = toBelCatalystActivity(entity)
return bel
def toBel(dbId):
''' Convert to BEL formats '''
entity = getEntityData(dbId)
print('toBel dbId: ', dbId)
type = entity['schemaClass']
if (type == 'Compartment'):
bel = toBelCompartment(entity)
elif type == 'EntityCompartment':
bel = toBelEntityCompartment(entity)
elif type == 'OtherEntity':
bel = toBelOtherEntity(entity)
elif type == 'Polymer':
bel = toBelPolymer(entity)
elif type == 'Complex':
bel = toBelComplexComponents(
entity) # alternative version - toBelComplexNamed()
elif type == 'SimpleEntity':
bel = toBelSimpleEntity(entity)
elif type == 'EntityWithAccessionedSequence':
bel = toBelEntityWithAccessionedSequence(entity)
elif 'Set' in type:
bel = toBelSets(entity)
elif 'GenomeEncodedEntity' in type:
bel = toBelGenomeEncodedEntity(entity)
elif 'CatalystActivity' in type:
bel = toBelCatalystActivity(entity)
return bel
def buildBelEvidences(reactionList, belversion, pathways=None):
''' Load reactions and build BEL Evidences'''
rxnUrlTpl = 'http://www.reactome.org/PathwayBrowser/#'
evidences = []
bad_namespaces_evidences = [] # collect bad evidences
# indexCnt = 0
for rxnId, rxnName in reactionList:
print('rxnId: ', rxnId)
# indexCnt += 1
# if indexCnt > 2:
# break
# Process Annotation information
rxnData = getEntityData(rxnId)
if 'stableIdentifier' not in rxnData:
continue
stableId = rxnData['stableIdentifier']['displayName']
rxnUrl = '{}{}'.format(rxnUrlTpl, stableId)
rxnName = escapeBelString(rxnData['displayName'])
rxnType = rxnData['schemaClass']
# Todo collect all compartments and annotate
compartment = 'Unknown'
if 'compartment' in rxnData:
compartment = rxnData['compartment'][0]['displayName']
rxnAuthor = rxnDate = None
if 'created' in rxnData:
try:
matches = re.search(r'(.*?),\s+(\d{4,4}-\d{2,2}-\d{2,2})',
rxnData['created']['displayName'])
if matches:
rxnAuthor = matches.group(1)
rxnDate = matches.group(2)
except:
log.info(
'Rxn - cannot find created date and author in object: {}'.
format(rxnId))
if rxnDate and rxnAuthor:
citation = '{{"Online Resource", "{}", "{}", "{}", "{}"}}'.format(
rxnName, rxnUrl, rxnDate, rxnAuthor)
else:
citation = '{{"Online Resource", "{}", "{}"}}'.format(
rxnName, rxnUrl)
evidence = {
'name': rxnName,
'rxnId': rxnId, # TODO remove after debugging
'rxnType': rxnType,
'compartment': compartment,
'species': rxnData['speciesName'],
'species_tax_id':
convertSpeciesNameToTaxId(rxnData['speciesName']),
'summary_text': rxnName,
'citation': citation,
}
# Process BEL Statement
catalysts, inputs, outputs = [], [], []
if 'catalystActivity' in rxnData:
for catalyst in rxnData['catalystActivity']:
catalysts.append(toBel(catalyst['dbId']))
# print('Catalyst: {}'.format(catalyst['dbId']))
if 'input' in rxnData:
for input in dedup(rxnData['input']):
inputs.append(toBel(input['dbId']))
if 'output' in rxnData:
for output in dedup(rxnData['output']):
outputs.append(toBel(output['dbId']))
print('Catalysts ', catalysts)
print('Inputs ', inputs)
print('Outputs ', outputs)
print('\n')
statements = buildStatements(catalysts, inputs, outputs)
evidence['statements'] = dedupList(statements)
bad_namespace_flag = False
for statement in statements:
if 'ENSEMBL' in statement or 'EMBL' in statement:
bad_namespace_flag = True
# with open('tmp_evidences.json', 'a') as f:
# json.dump(evidence, f, indent=4)
# f.write('\n\n')
if bad_namespace_flag:
bad_namespaces_evidences.append(copy.deepcopy(evidence))
else:
evidences.append(copy.deepcopy(evidence))
belscript = render_template(template_filename,
evidences,
pathways=pathways)
fn = 'reactome.bels'
if belversion == '2':
fn += '2'
with open(fn, 'w') as f:
f.write(belscript)
import json
with open('bad_evidences.json', 'w') as f:
json.dump(bad_namespaces_evidences, f, indent=4)
def buildBelEvidences(reactionList, belversion, pathways=None):
''' Load reactions and build BEL Evidences'''
rxnUrlTpl = 'http://www.reactome.org/PathwayBrowser/#'
evidences = []
bad_namespaces_evidences = [] # collect bad evidences
# indexCnt = 0
for rxnId, rxnName in reactionList:
print('rxnId: ', rxnId)
# indexCnt += 1
# if indexCnt > 2:
# break
# Process Annotation information
rxnData = getEntityData(rxnId)
if 'stableIdentifier' not in rxnData:
continue
stableId = rxnData['stableIdentifier']['displayName']
rxnUrl = '{}{}'.format(rxnUrlTpl, stableId)
rxnName = escapeBelString(rxnData['displayName'])
rxnType = rxnData['schemaClass']
# Todo collect all compartments and annotate
compartment = 'Unknown'
if 'compartment' in rxnData:
compartment = rxnData['compartment'][0]['displayName']
rxnAuthor = rxnDate = None
if 'created' in rxnData:
try:
matches = re.search(r'(.*?),\s+(\d{4,4}-\d{2,2}-\d{2,2})', rxnData['created']['displayName'])
if matches:
rxnAuthor = matches.group(1)
rxnDate = matches.group(2)
except:
log.info('Rxn - cannot find created date and author in object: {}'.format(rxnId))
if rxnDate and rxnAuthor:
citation = '{{"Online Resource", "{}", "{}", "{}", "{}"}}'.format(rxnName, rxnUrl, rxnDate, rxnAuthor)
else:
citation = '{{"Online Resource", "{}", "{}"}}'.format(rxnName, rxnUrl)
evidence = {
'name': rxnName,
'rxnId': rxnId, # TODO remove after debugging
'rxnType': rxnType,
'compartment': compartment,
'species': rxnData['speciesName'],
'species_tax_id': convertSpeciesNameToTaxId(rxnData['speciesName']),
'summary_text': rxnName,
'citation': citation,
}
# Process BEL Statement
catalysts, inputs, outputs = [], [], []
if 'catalystActivity' in rxnData:
for catalyst in rxnData['catalystActivity']:
catalysts.append(toBel(catalyst['dbId']))
# print('Catalyst: {}'.format(catalyst['dbId']))
if 'input' in rxnData:
for input in dedup(rxnData['input']):
inputs.append(toBel(input['dbId']))
if 'output' in rxnData:
for output in dedup(rxnData['output']):
outputs.append(toBel(output['dbId']))
print('Catalysts ', catalysts)
print('Inputs ', inputs)
print('Outputs ', outputs)
print('\n')
statements = buildStatements(catalysts, inputs, outputs)
evidence['statements'] = dedupList(statements)
bad_namespace_flag = False
for statement in statements:
if 'ENSEMBL' in statement or 'EMBL' in statement:
bad_namespace_flag = True
# with open('tmp_evidences.json', 'a') as f:
# json.dump(evidence, f, indent=4)
# f.write('\n\n')
if bad_namespace_flag:
bad_namespaces_evidences.append(copy.deepcopy(evidence))
else:
evidences.append(copy.deepcopy(evidence))
belscript = render_template(template_filename, evidences, pathways=pathways)
fn = 'reactome.bels'
if belversion == '2':
fn += '2'
with open(fn, 'w') as f:
f.write(belscript)
import json
with open('bad_evidences.json', 'w') as f:
json.dump(bad_namespaces_evidences, f, indent=4)
