def getSBMLFromBiomodelsURN(urn):
"""
Get the SBML from a given BioModels URN.
Searches for a BioModels identifier in the given urn and retrieves the SBML from biomodels.
:param urn:
:type urn:
:return: SBML string fro given model urn
:rtype: str
"""
pattern = "((BIOMD|MODEL)\d{10})|(BMID\d{12})"
match = re.search(pattern, urn)
mid = match.group(0)
try:
# read SBML via bioservices
import bioservices
biomodels = bioservices.BioModels()
sbml = biomodels.getModelSBMLById(mid)
sbml = sbml.encode('utf8')
except ImportError:
# if there are issues with bioservice import, do a workaround
# see https://github.com/sys-bio/tellurium/issues/125
import httplib
conn = httplib.HTTPConnection("www.ebi.ac.uk")
conn.request("GET", "/biomodels-main/download?mid=" + mid)
r1 = conn.getresponse()
# print(r1.status, r1.reason)
sbml = r1.read()
conn.close()
return str(sbml)
def __init__(self, debug=False):
""" Creates and displays the search form. """
self.debug = debug
self.s = bioservices.BioModels()
self.ch = bioservices.ChEBI()
# Define widgets
# <Search>
self.wSearchTerm = w.Text(description='Search biomodels by species:',
value="CHEBI:17925")
self.wSearchTerm.on_submit(self.searchChebi)
self.wSearchButton = w.Button(description='Search')
self.wSearchButton.on_click(self.searchChebi)
self.wSearchChebi = w.HBox(
children=[self.wSearchTerm, self.wSearchButton])
self.wSelectChebis = w.Select(description='Matching ChEBI:',
width='600px',
height='250px')
# FIXME: update the deprecated functions
self.wSelectChebis.on_trait_change(self.selectChebi)
self.wSelectModels = w.Select(description='Matching BioModels:',
width='200px')
self.wSelectModels.on_trait_change(self.selectedModel)
# <Model>
self.wModelId = w.Text(description='Model ID:',
value="No model selected")
self.wModelCode = w.Text(description='Install Code:')
self.wModelImport = w.Text(description='Import module code:')
self.wModelSbml = w.Textarea(description='Model SBML:',
width='800px',
height='300px')
# <Container>
self.wModel = w.FlexBox(children=[
self.wModelId, self.wModelCode, self.wSelectModels,
self.wModelImport, self.wModelSbml
])
for ch in self.wModel.children:
ch.font_family = 'monospace'
ch.color = '#AAAAAA'
ch.background_color = 'black'
self.wContainer = w.FlexBox(
children=[self.wSearchChebi, self.wSelectChebis, self.wModel])
# display the widgets
display(self.wContainer)
# clear notebook output
clear_output()
def compressionDistroAnalysis(directory):
'''
analyses the model compression distribution
(number of molecule types)/total species
acccording to what the model does
'''
with open('{0}/sortedD.dump'.format(directory),'rb') as evaluationFile:
ev1 = pickle.load(evaluationFile)
ev2 = []
for x in ev1:
try:
ev2.append([x['index'],x['nreactions'],x['nspecies'],x['compression']])
except:
continue
number,rulesLength,speciesLength,evaluation2 = zip(*ev2)
trueRatio = []
#with open('ratomization.dump','rb') as f:
# ratomizationDict = pickle.load(f)
for x,z,w in zip(rulesLength,number,evaluation2):
if x>=0:
trueRatio.append([z,1-w])
bio = bioservices.BioModels()
modelAnnotationArray = []
for element,compression in trueRatio:
try:
info = bio.getSimpleModelsByIds('BIOMD%010i' % element)
parsedInfo = bioservices.Service('name').easyXML(info)
modelName = removeTags(str(parsedInfo['modelname'][0]))
modelName = modelName.split(' - ')
if len(modelName) > 1:
modelName = modelName[1]
else:
modelName = modelName[0].split('_')
if len(modelName) > 1:
modelName = ' '.join(modelName[1:])
else:
modelName = modelName[0]
modelAnnotationArray.append([element,compression,modelName])
print modelAnnotationArray[-1]
except:
print parsedInfo
break
with open('compressionAnnotation.dump','wb') as f:
pickle.dump(modelAnnotationArray,f)
def download_models():
bio = bioservices.BioModels()
print 'The number of models in biomodels right now is {}'.format(len(bio))
m = bio.getAllCuratedModelsId()
print 'The number of curated models in biomodels is: {}'.format(len(m))
model_dct = {}
skipped = 0
for i in m:
os.chdir(d)
dire = os.path.join(d, i)
if os.path.isdir(dire) == False:
os.mkdir(dire)
else:
skipped += 1
continue
models_to_skip = ['BIOMD0000000241', 'BIOMD0000000148']
if i in models_to_skip:
'''
This file is broken and doesn't simulate with CopasiSE
'''
continue
try:
name = bio.getModelNameById(i)
strings = '\[\]_\{\}'
name = re.sub(strings, '_', name)
model_dct[name] = bio.getModelSBMLById(i)
print 'downloading {}:\t{}'.format(i, name.encode('utf8'))
fle = os.path.join(dire, name + '.xml')
if os.path.isfile(fle) != True:
with open(fle, 'w') as f:
f.write(model_dct[name].encode('utf8'))
time.sleep(0.25)
except:
continue
print 'You have downloaded {} out of {} models'.format(
len(model_dct.keys()), len(m))
print 'you have skipped {} models because you already have a folder for them'.format(
skipped)
df = pandas.DataFrame.from_dict(model_dct.keys())
xlsx = os.path.join(d, 'ModelsMap.xlsx')
df.to_excel(xlsx, index=True, header=True)
return df
def download_models(directory, percent=100, SKIP_ALREADY_DOWNLOADED=True):
'''
download curated models from biomodels curated section
args:
directory:
Name of directory to download models too
===============================================================
Returns:
df:
containing models
pickle:
save to file and contains models
'''
if percent > 100 or percent < 0:
raise TypeError('percent should be between 0 and 100')
try:
import bioservices
except ImportError:
## install bioservices if it doesn't already exist
##May need to do this with admin rights
os.system('pip install bioservices')
import bioservices
except WindowsError:
raise ImportError(
'Need bioservices module to download biomodels database. Use pip install bioservices with admin rights'
)
## create directory if not exist
if os.path.isdir(directory) != True:
os.makedirs(directory)
## change to directory
os.chdir(directory)
## get BioModels service
bio = bioservices.BioModels()
print 'The number of models in biomodels right now is {}'.format(len(bio))
model = bio.getAllCuratedModelsId()
print 'The number of curated models in biomodels is: {}'.format(len(model))
per = len(model) // 100.0 * percent
print 'You are about to download {} models'.format(per)
model_dct = {}
model_files = []
skipped = 0
if percent == 100:
models = model
else:
models = model[:int(per)]
for i in enumerate(models):
os.chdir(directory)
author = bio.getAuthorsByModelId(i[1])
author = RemoveNonAscii(author[0]).filter
dire = os.path.join(directory, i[1] + author)
# if SKIP_ALREADY_DOWNLOADED:
if os.path.isdir(dire) == False:
os.mkdir(dire)
else:
if SKIP_ALREADY_DOWNLOADED:
skipped += 1
continue
models_to_skip = ['BIOMD0000000241',
'BIOMD0000000148'] #these cause python to crash
if i[1] in models_to_skip:
'''
These file is broken and doesn't simulate with CopasiSE
'''
continue
try:
model_dct[author] = bio.getModelSBMLById(i[1])
print 'downloading model {} of {}: {}:\t{}'.format(
i[0], per, i[1], author.encode('utf8'))
fle = os.path.join(dire, author + '.xml')
print fle
if os.path.isfile(fle) != True:
with open(fle, 'w') as f:
f.write(model_dct[author].encode('utf8'))
time.sleep(0.25)
model_files.append(fle)
print 'saved to : {}'.format(fle)
except UnicodeEncodeError:
print 'model with author {} skipped as the name contains non-ascii characters'.format(
author)
continue
print 'You have downloaded {} out of {} models'.format(
len(model_dct.keys()), len(model))
print 'you have skipped {} models because you already have a folder for them'.format(
skipped)
df = pandas.DataFrame(model_files)
pickle_file = os.path.join(directory, 'BioModelsFilesPickle.pickle')
df.to_pickle(pickle_file)
return df