def __init__(self, configpath):
with open(configpath, 'r') as f:
self.config = yaml.load(f.read())
try:
self.nfsim = NFSim(
os.path.join(self.config['libpath'], 'libnfsim_c.so'))
except OSError:
print("Cannot open libnfsim_c.so. Please check libpath in "
"mcellr.yaml")
sys.exit(0)
def __init__(self, configpath):
self.config = {}
self.config['bionetgen'] = os.path.join(configpath, 'bng2', 'BNG2.pl')
self.config['mcell'] = os.path.join(configpath, 'mcell')
self.config['libpath'] = os.path.join(configpath, 'lib')
self.config['scriptpath'] = configpath
prefix = "lib"
if (sys.platform == 'linux') or (sys.platform == 'linux2'):
extension = "so"
elif (sys.platform == 'darwin'):
extension = "dylib"
elif (sys.platform == 'win32'):
extension = "dll"
else:
raise Exception("Unexpected platform: {0}".format(sys.platform))
libNFsim_path = os.path.join(self.config['libpath'],
'{0}NFsim.{1}'.format(prefix, extension))
libnfsim_c_path = os.path.join(
self.config['libpath'], '{0}nfsim_c.{1}'.format(prefix, extension))
if not os.path.exists(libNFsim_path):
# try the build directory paths
libNFsim_path = os.path.join(
self.config['libpath'], '..', 'nfsim',
'{0}NFsim.{1}'.format(prefix, extension))
if not os.path.exists(libnfsim_c_path):
# try the cygwin variant
libnfsim_c_path = os.path.join(
self.config['libpath'], '..', 'nfsimCInterface',
'{0}nfsim_c.{1}'.format(prefix, extension))
if not os.path.exists(libNFsim_path):
# default to system search using LD_LIBRARY_PATH on Unix or PATH on Windows
libNFsim_path = '{0}NFsim.{1}'.format(prefix, extension)
if not os.path.exists(libnfsim_c_path):
# default to system search
libnfsim_c_path = '{0}nfsim_c.{1}'.format(prefix, extension)
print("Loading libs from " + libNFsim_path + " and " +
libnfsim_c_path + ".")
self.nfsim = NFSim(libnfsim_c_path, libNFsim_path=libNFsim_path)
class MDLR2MDL(object):
'''
given an mdlr definition this script will get a set of mdl files compatible
with an nfsim bng-xml definition
'''
def __init__(self, configpath):
with open(configpath, 'r') as f:
self.config = yaml.load(f.read())
try:
self.nfsim = NFSim(
os.path.join(self.config['libpath'], 'libnfsim_c.so'))
except OSError:
print("Cannot open libnfsim_c.so. Please check libpath in "
"mcellr.yaml")
sys.exit(0)
def process_mdlr(self, mdlrPath):
'''
main method. extracts species definition, creates bng-xml, creates mdl
definitions
'''
try:
nauty_dict = self.xml2hnauty_species_definitions(mdlrPath)
except OSError:
print("Cannot open BNG2.pl. Please check bionetgen in mcellr.yaml")
sys.exit(0)
# append extended bng-xml to the bng-xml definition (the one that
# doesn't include seed information)
bngxmlestr = writeBXe.merge_bxbxe(namespace.input + '_total.xml',
namespace.input + '_extended.xml')
with open(mdlrPath + '_total.xml', 'w') as f:
f.write(bngxmlestr)
xmlspec = read_bngxml.parseFullXML(namespace.input + '.xml')
# write out the equivalent plain mdl stuffs
mdl_dict = write_mdl.construct_mcell(xmlspec, namespace.input,
final_name.split(os.sep)[-1],
nauty_dict)
write_mdl.write_mdl(mdl_dict, final_name)
def tokenize_seed_elements(self, seed):
# extract species names
seedKeys = re.findall(
'concentration="[0-9a-zA-Z_]+" name="[_0-9a-zA-Z@:(~!),.]+"', seed)
seedKeys = [
re.sub(
'concentration="([0-9a-zA-Z_]+)" name="([_0-9a-zA-Z@:(~!),.]+)"',
'\g<2>;\g<1>', x) for x in seedKeys
]
seedList = seed.split('</Species>')
seedList = [x.strip() for x in seedList if x != '']
seedList = [x + '</Species>' for x in seedList]
seedList = [re.sub('"S[0-9]+"', "S1", x) for x in seedList]
seedList = [
re.sub('concentration="[0-9a-zA-Z_]+"', 'concentration="1"', x)
for x in seedList
]
seedList = [
'<Model><ListOfSpecies>{0}</ListOfSpecies></Model>'.format(x)
for x in seedList
]
# seed_dict = {x:y for x, y in zip(seedKeys, seedList)}
seed_dict = {
x.split(';')[0]: y
for x, y in zip(seedKeys, seedList) if x.split(';')[1] != '0'
}
# print '---', seed_dict.keys()
return seed_dict
def get_names_from_definition_string(self, defStr):
species_names = re.findall('[0-9a-zA-Z_]+\(', defStr)
return [x[:-1] for x in species_names]
def xml2hnauty_species_definitions(self, inputMDLRFile):
"""
Temporary function for translating xml bng definitions to nauty species
definition strings
it call the nfsim library to get the list of possible complexes in the
system, however the function right now returns the species in question
+ all molecule types (if we are sending a lone molecule tye as
initialization it still returns all molecule types), which means the
list requires filterinng. and the filtering is not pretty
How to solve: make it so that nfsim returns a more sensible complex
list (filtering out unrelated molecule types) or create a nauty label
creation mechanism locally
"""
command = [
self.config['bionetgen'], '-xml', '-check', inputMDLRFile + '.bngl'
]
output_dir = os.path.dirname(inputMDLRFile)
if output_dir:
command.extend(['--outdir', output_dir])
# get a bng-xml file
call(command)
# extract seed species defition
seed, rest = split_bngxml.extractSeedBNG(inputMDLRFile + '.xml')
# store xml with non-seed sections and load up nfsim library
with open(namespace.input + '_total.xml', 'w') as f:
f.write(rest)
# load up nfsim library
self.nfsim.init_nfsim(namespace.input + '_total.xml', 0)
# remove encapsulating tags
seed = seed[30:-30]
# get the seed species definitions as a list
seed_dict = self.tokenize_seed_elements(seed)
nauty_dict = {}
for seed in seed_dict:
# initialize nfsim with each species definition and get back a
# dirty list where one of the entries is the one we want
#
# XXX: i think i've solved it on the nfsim side, double check
tmpList = self.get_nauty_string(seed_dict[seed])
# and now filter it out...
# get species names from species definition string
species_names = self.get_names_from_definition_string(seed)
nauty_dict[seed] = [
x for x in tmpList if all(y in x for y in species_names)
][0]
return nauty_dict
def get_nauty_string(self, xmlSpeciesDefinition):
self.nfsim.reset_system()
self.nfsim.init_system_xml(xmlSpeciesDefinition)
result = self.nfsim.querySystemStatus("complex")
return result
def __init__(self, configpath):
with open(configpath, 'r') as f:
self.config = yaml.load(f.read())
self.nfsim = NFSim(os.path.join(self.config['libpath'], 'libnfsim_c.so'))
class MDLR2MDL(object):
'''
given an mdlr definition this script will get a set of mdl files compatible with an nfsim bng-xml definition
'''
def __init__(self, configpath):
with open(configpath, 'r') as f:
self.config = yaml.load(f.read())
self.nfsim = NFSim(os.path.join(self.config['libpath'], 'libnfsim_c.so'))
def processMDLR(self, mdlrPath):
'''
main method. extracts species definition, creates bng-xml, creates mdl definitions
'''
nautyDict = self.xml2HNautySpeciesDefinitions(mdlrPath)
#append extended bng-xml to the bng-xml definition (the one that doesn't include seed information)
bngxmlestr = writeBXe.mergeBXBXe(namespace.input + '_total.xml', namespace.input + '_extended.xml')
with open(mdlrPath + '_total.xml', 'w') as f:
f.write(bngxmlestr)
xmlspec = readBNGXML.parseFullXML(namespace.input + '.xml')
# write out the equivalent plain mdl stuffs
mdlDict = writeMDL.constructMCell(xmlspec, namespace.input, finalName.split(os.sep)[-1], nautyDict)
writeMDL.writeMDL(mdlDict, finalName)
def tokenizeSeedElements(self, seed):
#extract species names
seedKeys = re.findall('concentration="[0-9a-zA-Z_]+" name="[_0-9a-zA-Z@:(~!),.]+"', seed)
seedKeys = [re.sub('concentration="([0-9a-zA-Z_]+)" name="([_0-9a-zA-Z@:(~!),.]+)"', '\g<2>;\g<1>', x) for x in seedKeys]
seedList = seed.split('</Species>')
seedList = [x.strip() for x in seedList if x != '']
seedList = [x + '</Species>' for x in seedList]
seedList = [re.sub('"S[0-9]+"', "S1", x) for x in seedList]
seedList = [re.sub('concentration="[0-9a-zA-Z_]+"', 'concentration="1"', x) for x in seedList]
seedList = ['<Model><ListOfSpecies>{0}</ListOfSpecies></Model>'.format(x) for x in seedList]
#seedDict = {x:y for x,y in zip(seedKeys,seedList)}
seedDict = {x.split(';')[0]:y for x, y in zip(seedKeys, seedList) if x.split(';')[1] != '0'}
#print '---', seedDict.keys()
return seedDict
def getNamesFromDefinitionString(self, defStr):
speciesNames = re.findall('[0-9a-zA-Z_]+\(',defStr)
return [x[:-1] for x in speciesNames]
def xml2HNautySpeciesDefinitions(self, inputMDLRFile):
"""
Temporary function for translating xml bng definitions to nautty species definition strings
it call the nfsim library to get the list of possible complexes in the system, however the function right now
returns the species in question + all molecule types (if we are sending a lone molecule tye as initialization
it still returns all molecule types), which means the list requires filterinng. and the filtering
is not pretty
How to solve: make it so that nfsim returns a more sensible complex list (filtering out unrelated molecule types) or
create a nauty label creation mechanism locally
"""
#get a bng-xml file
call([self.config['bionetgen'], '-xml', '-check', inputMDLRFile + '.bngl'])
#extract seed species defition
seed, rest = splitBNGXML.extractSeedBNG(inputMDLRFile + '.xml')
#store xml with non-seed sections and load up nfsim library
with open(namespace.input + '_total.xml', 'w') as f:
f.write(rest)
#load up nfsim library
self.nfsim.initNFsim(namespace.input + '_total.xml', 0)
# remove encapsulating tags
seed = seed[30:-30]
#get the seed species definitions as a list
seedDict = self.tokenizeSeedElements(seed)
nautyDict = {}
for seed in seedDict:
#initialize nfsim with each species definition and get back a dirty list where one of the entries is the one we want
#XXX: i think i've solved it on the nfsim side, double check
tmpList = self.getNautyString(seedDict[seed])
#and now filter it out...
#get species names from species definition string
speciesNames = self.getNamesFromDefinitionString(seed)
nautyDict[seed] = [x for x in tmpList if all(y in x for y in speciesNames)][0]
return nautyDict
def getNautyString(self, xmlSpeciesDefinition):
self.nfsim.resetSystem()
self.nfsim.initSystemXML(xmlSpeciesDefinition)
result = self.nfsim.querySystemStatus("complex")
return result
class MDLR2MDL(object):
'''
given an mdlr definition this script will get a set of mdl files compatible
with an nfsim bng-xml definition
'''
def __init__(self, configpath):
self.config = {}
self.config['bionetgen'] = os.path.join(configpath, 'bng2', 'BNG2.pl')
self.config['mcell'] = os.path.join(configpath, 'mcell')
self.config['libpath'] = os.path.join(configpath, 'lib')
self.config['scriptpath'] = configpath
prefix = "lib"
if (sys.platform == 'linux') or (sys.platform == 'linux2'):
extension = "so"
elif (sys.platform == 'darwin'):
extension = "dylib"
elif (sys.platform == 'win32'):
extension = "dll"
else:
raise Exception("Unexpected platform: {0}".format(sys.platform))
libNFsim_path = os.path.join(self.config['libpath'],
'{0}NFsim.{1}'.format(prefix, extension))
libnfsim_c_path = os.path.join(
self.config['libpath'], '{0}nfsim_c.{1}'.format(prefix, extension))
if not os.path.exists(libNFsim_path):
# try the build directory paths
libNFsim_path = os.path.join(
self.config['libpath'], '..', 'nfsim',
'{0}NFsim.{1}'.format(prefix, extension))
if not os.path.exists(libnfsim_c_path):
# try the cygwin variant
libnfsim_c_path = os.path.join(
self.config['libpath'], '..', 'nfsimCInterface',
'{0}nfsim_c.{1}'.format(prefix, extension))
print("Loading libs from " + libNFsim_path + " and " +
libnfsim_c_path + ".")
self.nfsim = NFSim(libnfsim_c_path, libNFsim_path=libNFsim_path)
def process_mdlr(self, mdlrPath):
'''
main method. extracts species definition, creates bng-xml, creates mdl
definitions
'''
try:
nauty_dict = self.xml2hnauty_species_definitions(mdlrPath)
except OSError:
print(
'Cannot open BNG2.pl. Please check BioNetGen is installed at: %s'
% (self.config['bionetgen']))
sys.exit(0)
# append extended bng-xml to the bng-xml definition (the one that
# doesn't include seed information)
bngxmlestr = writeBXe.merge_bxbxe(
namespace.input + '_rules.xml',
namespace.input + '_extended_bng.xml')
with open(mdlrPath + '_rules.xml', 'w') as f:
f.write(bngxmlestr)
xmlspec = read_bngxml.parseFullXML(namespace.input + '.xml')
# write out the equivalent plain mdl stuffs
mdl_dict = write_mdl.construct_mcell(xmlspec, namespace.input,
final_name.split(os.sep)[-1],
nauty_dict)
write_mdl.write_mdl(mdl_dict, final_name)
def tokenize_seed_elements(self, seed):
# extract species names
seedKeys = re.findall(
'concentration="[0-9a-zA-Z_]+" name="[_0-9a-zA-Z@:(~!),.]+"', seed)
seedKeys = [
re.sub(
'concentration="([0-9a-zA-Z_]+)" name="([_0-9a-zA-Z@:(~!),.]+)"',
'\g<2>;\g<1>', x) for x in seedKeys
]
seedList = seed.split('</Species>')
seedList = [x.strip() for x in seedList if x != '']
seedList = [x + '</Species>' for x in seedList]
# Jose's code here seems to go overboard and also rename components named S[0-9]+
# The intent seems to be to rename Species id's from "S[0-9]+" to S1, with double quotes dropped
#
# seedList = [re.sub('"S[0-9]+"', "S1", x) for x in seedList]
#
# Attempt to fix this by parsing the full context of 'Species id=...'
seedList = [
re.sub('Species id="S[0-9]+"', 'Species id=S1', x)
for x in seedList
]
seedList = [
re.sub('concentration="[0-9a-zA-Z_]+"', 'concentration="1"', x)
for x in seedList
]
seedList = [
'<Model><ListOfSpecies>{0}</ListOfSpecies></Model>'.format(x)
for x in seedList
]
# seed_dict = {x:y for x, y in zip(seedKeys, seedList)}
seed_dict = {
x.split(';')[0]: y
for x, y in zip(seedKeys, seedList) if x.split(';')[1] != '0'
}
# print '---', seed_dict.keys()
return seed_dict
def get_names_from_definition_string(self, defStr):
species_names = re.findall('[0-9a-zA-Z_]+\(', defStr)
return [x[:-1] for x in species_names]
def xml2hnauty_species_definitions(self, inputMDLRFile):
"""
Temporary function for translating xml bng definitions to nauty species
definition strings
it call the nfsim library to get the list of possible complexes in the
system, however the function right now returns the species in question
+ all molecule types (if we are sending a lone molecule tye as
initialization it still returns all molecule types), which means the
list requires filtering. and the filtering is not pretty
How to solve: make it so that nfsim returns a more sensible complex
list (filtering out unrelated molecule types) or create a nauty label
creation mechanism locally
"""
command = [
'perl', self.config['bionetgen'], '-xml', '-check',
inputMDLRFile + '.bngl'
]
output_dir = os.path.dirname(inputMDLRFile)
if output_dir:
command.extend(['--outdir', output_dir])
# get a bng-xml file
print("\n====> Running BioNetGen with explicit \"perl\": " +
" ".join(command) + "\n")
call(command)
# extract seed species definition
seed, rest = split_bngxml.extractSeedBNG(inputMDLRFile + '.xml')
# store xml with non-seed sections and load up nfsim library
print("\nStore xml with non-seed sections and load up nfsim library\n")
with open(namespace.input + '_rules.xml', 'w') as f:
f.write(rest)
# load up nfsim library
print("Initializing NFSim using: " + namespace.input + '_rules.xml')
self.nfsim.init_nfsim(namespace.input + '_rules.xml', 0)
# remove encapsulating tags
seed = seed[30:-30]
# get the seed species definitions as a list
# print(">>>>>>>>>> THE SEED LIST: \n", str(seed))
seed_dict = self.tokenize_seed_elements(seed)
nauty_dict = {}
# print(">>>>>>>>>> SEED DICT: \n", seed_dict)
for seed in seed_dict:
# initialize nfsim with each species definition and get back a
# dirty list where one of the entries is the one we want
#
# XXX: i think i've solved it on the nfsim side, double check
tmpList = self.get_nauty_string(seed_dict[seed])
# print('>>>>>>>> SEED_DICT[SEED]: ' + str(seed_dict[seed]))
# print('>>>>>>>> tmpList: ' + str(tmpList))
# and now filter it out...
# get species names from species definition string
species_names = self.get_names_from_definition_string(seed)
nauty_dict[seed] = [
x for x in tmpList if all(y in x for y in species_names)
][0]
return nauty_dict
def get_nauty_string(self, xmlSpeciesDefinition):
self.nfsim.reset_system()
self.nfsim.init_system_xml(xmlSpeciesDefinition)
result = self.nfsim.querySystemStatus("complex")
return result
