def biomass(self, biomass_file, readquiet):
#read and define biomass equation...
if not readquiet:
print 'biomass from', biomass_file
biomass_equation = [[], []]
file = open(biomass_file)
while True:
line = file.readline()
if line == '': break
line = line.rstrip()
if line == '': continue
if line[0] == '#': continue
col = line.split('\t')
[rawmet, coef, side] = col[0:3]
#convert metabolites from forms like 'leu-L[c]' to 'M_leu_DASH_L_c'...
met = eq_current.convert_metabolite_ext2int(rawmet)
if 'reactant' in side:
biomass_equation[0].append((met, coef))
self.NOTSOURCES.append(met)
elif 'product' in side:
biomass_equation[1].append((met, coef))
self.NOTSOURCES.append(met)
else:
print 'Warning: biomass component skipped because cannot be idenified as reactant or product-->', line
#add biomass equation to model, and set this as the objective...
cb.add_reaction(self, 'R_biomass_target', 'BiomassRxn', False, {'SUBSYSTEM: BiomassObjective':1, 'EC: .':1}, biomass_equation)
cb.set_objective(self, 'Maximize', 'R_biomass_target')
#cb.set_constraint(self, 'R_biomass_target', '0.0', str(float(self.VMAX)) )
if not readquiet:
print eq_current.makestring(biomass_equation, False)
print
def deletions (self, level):
"Delete all genes, proteins, or reactions, one at a time."
cb.solve(self, verbose=False)
fullstatus, fullobjectivevalue = self.STATUS, self.OBJECTIVE_VALUE
#item is a gene, protein, or reaction; level is self.GENES, self.PROTS, or self.REACTS
lethals = {}
for item in level:
#delete item
level[item] = 0
#calculate consequences of deleted item according to boolean rules (i.e., which reactions are eliminated?)
deletedrxns = cb.calc(self)
#now constrain each reaction that is deleted by the change to have zero flux, attempt fba
for r in deletedrxns:
name, rev, notes, equation = self.REACTIONS[r]
if rev:
lbound = '-' + self.VMAX
else:
lbound = '0'
default_lbound, default_ubound = self.CONSTRAINTS.get(r, (lbound, self.VMAX))
cb.set_constraint(self, r, 0, 0)
cb.solve(self, verbose=False)
#if status != OPTIMAL or objective value is < 25% of 'wild type', print item, reaction, and results
if (not self.STATUS == 'OPTIMAL') or (float(self.OBJECTIVE_VALUE) < 0.25 * float(fullobjectivevalue)):
reactionequation = eq_current.makestring(equation, rev)
print item + '\t' + r + '\t' + reactionequation + '\t' + self.STATUS + '\t' + str(self.OBJECTIVE_VALUE)
lethals[item] = 1
#reset reaction constraints to default
cb.set_constraint(self, r, default_lbound, default_ubound)
#make item (gene, protein, ...) available again
level[item] = 1
return lethals
def ddeletions (self): "Double deletions at the reaction level." cb.solve(self, verbose=False) essential_reactions = cb.deletion_testing(self) fullstatus, fullobjectivevalue = self.STATUS, self.OBJECTIVE_VALUE for i, r in enumerate(self.REACTIONS.keys()): if not r == self.OBJECTIVE[1] and not r in essential_reactions and not 'R_ESC' in r and not 'R_SRC' in r: name, rev, notes, equation = self.REACTIONS[r] if rev: lbound = '-' + self.VMAX else: lbound = '0' default_lbound, default_ubound = self.CONSTRAINTS.get(r, (lbound, self.VMAX)) cb.set_constraint(self, r, 0, 0) for j, r2 in enumerate(self.REACTIONS.keys()[i+1:]): if not r2 == self.OBJECTIVE[1] and not r2 in essential_reactions and not 'R_ESC' in r2 and not 'R_SRC' in r2: name2, rev2, notes2, equation2 = self.REACTIONS[r2] if rev2: lbound2 = '-' + self.VMAX else: lbound2 = '0' default_lbound2, default_ubound2 = self.CONSTRAINTS.get(r2, (lbound2, self.VMAX)) cb.set_constraint(self, r2, 0, 0) cb.solve(self, verbose=False) if (not self.STATUS == 'OPTIMAL') or (float(self.OBJECTIVE_VALUE) < 0.25 * float(fullobjectivevalue)): reactionequation1 = eq_current.makestring(equation, rev) reactionequation2 = eq_current.makestring(equation2, rev2) print self.STATUS + '\t' + str(self.OBJECTIVE_VALUE) print r + '\t' + reactionequation1 print r2 + '\t' + reactionequation2 print cb.set_constraint(self, r2, default_lbound2, default_ubound2) cb.set_constraint(self, r, default_lbound, default_ubound)
def add_reaction (self, ID, name, rev, notes, equation):
"Add a new reaction into the model. Example: m.make_reaction('R_ss_biomass', 'ssa biomass', 'false', {'CONFIDENCE: 1':1, 'SUBSYSTEM: biomass':1, 'GPR: ':1, 'EC Number: ':1}, [[('M_atp_c', '1')],[('M_adp_c', '1'), ('M_pi_c', '1')]])"
if ID in self.REACTIONS:
print ID, 'already in REACTIONS'
else:
self.REACTIONS[ID] = (name, rev, notes, equation)
if ID in DISCREPANCIES:
warning_equation = eq_current.makestring(equation, rev)
#print ID, 'discrepant across models. Using:', warning_equation
for species, coef in (equation[0] + equation[1]):
#add compartment if necessary
compartment, outside = abbrev2compartment[species[-1:]]
if not compartment in self.COMPARTMENTS:
cb.add_compartment(self, compartment, outside)
#add species if necessary
if not species in self.SPECIES:
name, db_compartment, charge, boundaryCondition = '.', '.', '.', 'false'
cb.add_species(self, species, name, compartment, charge, boundaryCondition)
model.scoringExpData(exprfile)
model.webScraping(organism)
calls[x] = makeCallsDict(model.REACTIONS, model.hirxnset, model.lowrxnset)
fluxes[x] = deepcopy(model.REACTION2FLUXVALUE)
# reset default constraints:
# for c in constraints:
# (lb, ub) = defaultConstraints[c[0]]
# model.set_constraint(c[0], lb, ub)
# model.print_flux_xls( fluxes, 'Ruppins_SpaceFlight_05_09_2012.xls', calls )
cache = {}
# for reaction in model.varkeys():
for reaction in model.REACTIONS:
name, reversible, notes, equation = model.REACTIONS[reaction]
reactionequation = eq_current.makestring(equation, reversible)
# search for any ec numbers and pathways in reaction notes; it IS possible for there to be > 1 ec or pathway for a given reaction
confidence, gpr = "?", "?"
holder = {"pathways": {}, "ecs": {}}
ref, prr = ".", "."
for note in notes:
if "SUBSYSTEM: " in note:
holder["pathways"][note[11:]] = 1
if "EC: " in note:
holder["ecs"][note[4:]] = 1
# print note
if "SCORES: " in note:
print "THIS IS THE SCORE", notes["SCORES: "]
Avg = notes["SCORES: "] + Avg
count += 1
if "CONFIDENCE: " in note:
def list_reactions (self, out=False, showfluxvalues=True):
"Prints a list of reactions from current model, organized by path, then ecnumber. Arguments are out=<fn>, showfluxvalues=<True/False>. Defaults are False, True."
cache = {}
for reaction in self.REACTIONS:
name, reversible, notes, equation = self.REACTIONS[reaction]
reactionequation = eq_current.makestring(equation, reversible)
#search for any ec numbers and pathways in reaction notes; it IS possible for there to be > 1 ec or pathway for a given reaction
confidence, gpr = '?', '?'
holder = {'pathways':{}, 'ecs':{}}
ref, prr = '.', '.'
for note in notes:
if 'SUBSYSTEM: ' in note:
holder['pathways'][note[11:]] = 1
if 'EC: ' in note:
holder['ecs'][note[4:]] = 1
if 'CONFIDENCE: ' in note:
confidence = note[12:]
if 'GPR: ' in note:
gpr = note[5:]
if 'Protein_reaction_relation: ' in note:
prr = note[note.find(' == ') + 4:]
if 'PMID: ' in note: # and not 'review' in note and not 'related_organism' in note
ref = ref + note.split(',')[0][6:] + ' '
if not ref == '.':
ref = 'PMIDs: ' + ref[1:-1]
#construct grr
grr = prr
for i in prr.split():
if '(' in i:
i = i[1:]
if ')' in i:
i = i[:-1]
if i in self.PROTEIN2GENE:
grr = grr.replace(i, self.PROTEIN2GENE[i])
#add to cache for printing...
for pathwayname in holder['pathways']:
for ec in holder['ecs']:
if not pathwayname in cache:
cache[pathwayname] = {}
if not ec in cache[pathwayname]:
cache[pathwayname][ec] = {}
#if this is printing the results of 'solve', then get flux activities...
#reactionID name rev pathway ec equation confidence gpr
if showfluxvalues and self.REACTION2FLUXVALUE.get(reaction, '.') != '0':
cache[pathwayname][ec][ ('\t').join((reaction, name, str(reversible), pathwayname, ec, self.REACTION2FLUXVALUE.get(reaction, '.'), reactionequation)) ] = 1
if not showfluxvalues:
cache[pathwayname][ec][ ('\t').join((reaction, name, str(reversible), pathwayname, ec, reactionequation, prr, grr, ref)) ] = 1
paths = cache.keys()
paths.sort()
if out:
outfi = open(out, 'w')
for path in paths:
ecs = cache[path].keys()
ecs.sort()
for ec in ecs:
for r in cache[path][ec]:
print >>outfi, r
#print >>outfi, '\n'
else:
for path in paths:
ecs = cache[path].keys()
ecs.sort()
for ec in ecs:
for r in cache[path][ec]:
print r
def get_equation (self, id): "Given a reactionID, prettyprint the reaction equation." name, reversible, notes, equation = self.REACTIONS[id] reactionequation = eq_current.makestring(equation, reversible) return reactionequation
def data_integration(self,gammas = (-0.1,0.1),exprfile = args.exp,forceBiomass = True, verbose = False):
"""
Ruppin analysis of syw using the datasets from Tetu et al., 2009 about phosphate stress.
"""
gprfname = self.infile + '.gpr'
#gap_Exchanges = "gapSRC_ESC.txt"
def makeCallsDict( allRxns, hirxns, lorxns ):
calls = {}
for r in hirxns:
assert r in allRxns and r not in lorxns
for r in lorxns:
assert r in allRxns and r not in hirxns
for r in allRxns:
if r in hirxns:
calls[r] = "On"
elif r in lorxns:
calls[r] = "Off"
else:
calls[r] = "No call"
for r in allRxns:
assert not calls[r] == None
return calls
organism = raw_input("What is the organism's name in PubMed? ")
self.model = mg.gurobicb()
w2m.wil2metmodel(self.wilfname)
#self.model.build_from_textfiles(modelfile=self.modelname + ".reactions", biomassfile=self.modelname + ".biomass", sourcesfile=self.modelname+".sources", escapesfile=self.modelname +".escapes", exchangesfile="", gapexfile=gap_Exchanges)
self.model.build_from_textfiles(modelfile=self.modelname + ".reactions", biomassfile=self.modelname + ".biomass", sourcesfile=self.modelname+".sources", escapesfile=self.modelname +".escapes", exchangesfile="")
if verbose:
rq = False
else:
rq = True
self.model.gpr2(gprfname,readquiet=rq)
fluxes = {}
calls = {}
self.model.solve(verbose=verbose)
fluxes['fba'] = deepcopy(self.model.REACTION2FLUXVALUE)
self.model.ruppin(exprfile,gammas,forceBiomass=forceBiomass)
#self.model.scoringExpData(exprfile)
#self.model.webScrapping(organism)
calls[self.outprefix] = makeCallsDict(self.model.REACTIONS,self.model.hirxnset,self.model.lowrxnset)
fluxes[self.outprefix] = deepcopy(self.model.REACTION2FLUXVALUE)
cache = {}
for reaction in self.model.REACTIONS:
name, reversible, notes, equation = self.model.REACTIONS[reaction]
reactionequation = eq_current.makestring(equation, reversible)
confidence,gpr = '?','?'
holder = {'pathways':{}, 'ecs':{}}
ref, prr = '.', '.'
for note in notes:
if 'SUBSYSTEM: ' in note:
holder['pathways'][note[11:]] = 1
if 'EC: ' in note:
holder['ecs'][note[4:]] = 1
if 'SCORES: ' in note:
print ("THIS IS THE SCORE", notes['SCORES: '])
Avg = notes['SCORES: '] + Avg
count += 1
if 'CONFIDENCE: ' in note:
confidence = note[12:]
if 'GPR: ' in note:
gpr = note[5:]
if 'Protein_reaction_relation: ' in note:
prr = note[note.find(' == ') + 4:]
if 'PMID: ' in note: # and not 'review' in note and not 'related_organism' in note
ref = ref + note.split(',')[0][6:] + ' '
if not ref == '.':
ref = 'PMIDs: ' + ref[1:-1]
grr = prr
for i in prr.split():
if '(' in i:
i = i[1:]
if ')' in i:
i = i[:-1]
if i in self.model.PROTEIN2GENE:
grr = grr.replace(i, model.PROTEIN2GENE[i])
#add to cache for printing...
for pathwayname in holder['pathways']:
for ec in holder['ecs']:
if not pathwayname in cache:
cache[pathwayname] = {}
if not ec in cache[pathwayname]:
cache[pathwayname][ec] = {}
cache[pathwayname][ec][ ('\t').join((reaction, name, str(reversible), pathwayname, ec, fluxes['fba'][reaction], fluxes[self.outprefix][reaction], reactionequation)) ] = 1
paths = cache.keys()
paths.sort()
outfile = open(self.outprefix+"integrated.xls","w")
for path in paths:
ecs = cache[path].keys()
ecs.sort()
for ec in ecs:
for r in cache[path][ec]:
outfile.write("%s\n" % r)
#Average = float(Avg)/float(count)
#print ('\n\n', "Average Score of Model: ", Avg, '\n')
#print ("Average Score of Model per Gene: ", Average, '\n')
self.model.writeECfile(self.outprefix + "ec3.txt")
metfile = open(self.outprefix + "metabolites3.txt","w")
numspecies = 0
for species in self.model.SPECIES.keys():
if not species.endswith("_b"):
metfile.write("%s\n" % (species))
numspecies += 1
numreactions = 0
for reactions in self.model.REACTIONS.keys():
if (not reactions.startswith("R_SRC")) and (not reactions.startswith("R_ESC")) and (not reactions.startswith("R_EXCH")):
numreactions += 1
del self.model
return numspecies,numreactions
del self.model
