def __init__(self):
#Boolean controls
self.verbose = True
self.unityControl = False
self.useControlPenalty = False
self.simoControl = 1.0
self.simoLimit = 0.0
#data holders
self.parseEquations = ParseEquations()
self.parseEquations.identity = "I"
self.originalObjective = {}
self.newObjective = ''
self.primeControlLimits = {}
self.dualControlLimits = {}
#ControlLibrary Values
self.regulationLibrary= {}
self.controlPrefixMap = {}
self.libdata = []
self.rGeneMap = None
self.geneList = None
#static value defaults
self.controlMin = 0
self.controlMax = 1
self.minOObjective = 0.0
self.posControlBound = 1e3
self.negControlBound = -1e3
self.posfluxControlBound = 100
self.negfluxControlBound = -100
self.coeffecentDefault = 0.0
self.objectiveCoeffecent = 1.0
self.regCoeffecent = 1.0
self.delta = 1e-4
self.controlDelta = 1e-6
#Variable Names
self.pName = "" #prefix for prime variables
self.dName = "__lamda__" #dual variables for row equality
self.cName = "__y__" #prefix for binary control variable"
self.cdName = "__sigma__" #prefix for slack dual variable
self.cvName = "__n__" #prefix for analog control variable"
self.cnvName = "__nv__" #prefix for slack analog control variable"
#Relation names
self.primeRName = "c1_" #prime limits
self.dualRName = "c2_" #dual limits
self.equalityRName = "c3_" #prime dual equality relation
class ConstructRegulationOptimization:
'''
primal:
min Z=c'*v
st
S*v = 0
a <= v <= b
dual:
max Z_dual = u*nu - l*mu
st
S'*lamda - mu + nu = c'
mu >= 0
nu >= o
Equality condition:
Z = Z_dual
c'*v = -l*mu + u*nu
Control:
Z += a'*n
y = 0 : n = 0
y = 1 : n = v
primal control:
n > 0
Z += a'*n
S += v - n > 0
n + vMax*y < vMax
n + vMin*y > vMin
Dual Control:
S'*lambda + nd = a'
n=0 -> nd = free or n=v -> nd = 0
sigma - sMax*y < 0
sigma + sMin*y > 0
columns:
: prime Names
__lambda__" dual names
__y__; control name
__n__: controled value
__nv__: negative controled value
rows:
__c1__: prime contstraints
__c2__: dual constraints
__c3__: equality constraints
'''
def __init__(self):
#Boolean controls
self.verbose = True
self.unityControl = False
self.useControlPenalty = False
self.simoControl = 1.0
self.simoLimit = 0.0
#data holders
self.parseEquations = ParseEquations()
self.parseEquations.identity = "I"
self.originalObjective = {}
self.newObjective = ''
self.primeControlLimits = {}
self.dualControlLimits = {}
#ControlLibrary Values
self.regulationLibrary= {}
self.controlPrefixMap = {}
self.libdata = []
self.rGeneMap = None
self.geneList = None
#static value defaults
self.controlMin = 0
self.controlMax = 1
self.minOObjective = 0.0
self.posControlBound = 1e3
self.negControlBound = -1e3
self.posfluxControlBound = 100
self.negfluxControlBound = -100
self.coeffecentDefault = 0.0
self.objectiveCoeffecent = 1.0
self.regCoeffecent = 1.0
self.delta = 1e-4
self.controlDelta = 1e-6
#Variable Names
self.pName = "" #prefix for prime variables
self.dName = "__lamda__" #dual variables for row equality
self.cName = "__y__" #prefix for binary control variable"
self.cdName = "__sigma__" #prefix for slack dual variable
self.cvName = "__n__" #prefix for analog control variable"
self.cnvName = "__nv__" #prefix for slack analog control variable"
#Relation names
self.primeRName = "c1_" #prime limits
self.dualRName = "c2_" #dual limits
self.equalityRName = "c3_" #prime dual equality relation
def _scaleMap(self,m,s):
'''
mulitply values in map m by scalar s
@type m: dict
@type s: float
'''
result = {}
for key in m.keys():
value = m[key]
result[key] = value*s
return result
def _confirmControlTargets(self,libraryName,targets):
libraryControlNames = self.regulationLibrary[libraryName]
result = set(libraryControlNames).intersection(targets)
return result
def _addControlLibrary(self,name,values):
ivalues = {}
for (key,value) in values.items():
ivalue = float(value)
if abs(ivalue) > self.controlDelta:
ivalues[key] = ivalue
else:
continue
self.regulationLibrary[name] = ivalues
return None
def extendControl(self,constructor):
self.regulationLibrary.update(constructor.regulationLibrary)
self.controlPrefixMap.update(constructor.controlPrefixMap)
for lib in constructor.libdata:
if lib not in self.libdata:
self.libdata.append(lib)
if self.rGeneMap == None:
self.rGeneMap = {}
self.rGeneMap.update(constructor.rGeneMap)
return None
def getGeneTargetMap(self):
if self.rGeneMap == None:
return None
result = {}
for (r,gs) in self.rGeneMap.items():
for g in gs:
if g not in result.keys():
result[g] = set()
result[g].add(r)
return result
def addRegulationLibrary(self,name,prefix,value):
'''
@type name: string
@type prefix: string
@type value: dict
'''
if len(value) == 0:
return None
self.controlPrefixMap[name] = prefix
self.libdata.append((name,prefix))
self._addControlLibrary(name,value)
return None
def clearControlLibraries(self):
self.regulationLibrary= {}
self.controlPrefixMap = {}
self.libdata = []
return None
def addRegulationLibraries(self,values):
'''
@type values: [(name,prefix,dict)]
'''
for v in values:
name = v[0]
prefix = v[1]
value = v[2]
self.addRegulationLibrary(name,prefix,value)
return None
def setGeneMap(self,data):
'''
@type data: {string,set(string)}
'''
if data == None:
return None
self.rGeneMap = data
self.geneList = set()
for (key,values) in data.items():
self.geneList = self.geneList.union(set(values))
return None
def _getRegulationCoeffecent(self,lib,name):
'''
Retrieve coeffecent value for control variable
@type name: string
@rtype: double
'''
if lib not in self.regulationLibrary.keys():
pass
if lib in self.regulationLibrary.keys():
l = self.regulationLibrary[lib]
if name in l.keys():
value = l[name]
return value
return self.coeffecentDefault
def _getNonZeroRegulationNames(self,libName,names):
result = []
for name in names:
value = self._getRegulationCoeffecent(libName,name)
if value != 0:
result.append(name)
return result
def getGeneControlMap(self,libName,geneName):
result = {}
if libName not in self.regulationLibrary.keys():
return None
controlMap = self.regulationLibrary[libName]
for (key,value) in controlMap.items():
if key not in self.rGeneMap:
continue
iGeneControls = self.rGeneMap[key]
if geneName in iGeneControls:
result[key] = value
return result
def getLibraryGeneControlMap(self,libName):
result = {}
for geneName in self.geneList:
geneControlMap = self.getGeneControlMap(libName, geneName)
if geneControlMap != {}:
result[geneName] = geneControlMap
return result
def _primalDual(self,model,objective,prime=True,dual=True,equal=True):
'''
class PathwayTarget:
@type model: LinearModel
@rtype: LinearModel
Construct Core Linear Optimization Matrix with primal dual equality
row i and column j
Primal:
min Z = sum(j): [c(j)*v(j)]
ST.
sum(j) [S(ij)*v(j)] = 0 E i
l(j) < v(j) < u(j) E j
Hameltonian:
Max(lambda,mu,nu): Min(v): sum(j)c(j)*v(j) + sum(i)lambda(i)*S(ij)v(j) + mu(j)(l(j)-v(j)) + nu(j)(v(j)-u(j))
=> c'v + lambda*(Sv) + mu(l-v) + nu(v-u)
=> Max: mu*l - nu*u
ST
c + lambda*S + mu + nu = 0
Dual:
max Z_dual = l*mu - u*nu
ST.
S'*lambda - mu + nu = -c'
mu(j) > 0 E j
nu(j) > 0 E j
Equality condition:
Z = Z_dual
==> sum(j) [c(j)'*v(j)] = sum(j) [l(j)*mu(j) - u(j)*nu]
==> -c*v + mu*l - u*nu = 0
'''
pName = self.pName
dName = self.dName
result = LinearModel()
#Primal:
if prime:
#S*v
for (row, column, value) in model.getSparseMatrix():
result.addData(self.primeRName + row, pName + column, value)
#S*v <=> b
for row in model.getRowLimits().keys():
result.addRowLimit(self.primeRName + row, model.getRowLimit(row))
#u<=v<=l
for column in model.getColumnNames():
result.addColumnLimit(pName + column, model.getColumnLimit(column))
#Dual:
if dual:
#S'*lambda
for (row, column, value) in model.getSparseMatrix():
result.addData(self.dualRName + column, dName + row, value)
#lambda <=> +- inf
for row in model.getRowLimits().keys():
result.addColumnLimit(dName + row, (None,None))
#S'*lambda = -c'
for column in model.getColumnNames():
if column in objective.keys():
value = objective[column]
result.addRowLimit(self.dualRName + column, (-value, -value))
else:
result.addRowLimit(self.dualRName + column, (0.0, 0.0))
#S'*lambda - mu + nu = -c'
for column in model.getColumnLimits().keys():
(lower, upper) = model.getColumnLimit(column)
if lower != None:
result.addData(self.dualRName + column, '__mu__' + column, -1)
result.addColumnLimit('__mu__' + column, (0.0, None))
if upper != None:
result.addData(self.dualRName + column, '__nu__' + column, 1)
result.addColumnLimit('__nu__' + column, (0.0, None))
if equal:
#objective relation
#Z^prime = Z^dual
# cv = l*mu - u*nu = 0
# -cv + l*mu - u*nu = 0
for column in model.getColumnNames():
if column in objective.keys():
value = objective[column]
result.addData(self.equalityRName, pName + column, -value)
else:
result.addData(self.equalityRName, pName + column, 0)
for column in model.getColumnLimits().keys():
(lower, upper) = model.getColumnLimit(column)
if lower != None and lower != 0.0:
result.addData(self.equalityRName, '__mu__' + column, lower)
if upper != None and upper != 0.0:
result.addData(self.equalityRName, '__nu__' + column, -upper)
result.addRowLimit(self.equalityRName,(0,0))
objective = {}
objective[self.pName + self.newObjective] = -1
result.setObjective(objective)
return result
def _constructRegulationObjecive(self,model,names,lib,prefix,dir):
'''
Hybrid Objective
Z += C'v + A'n
Control of adjustment objective in hybrid objective
if y = 1: n = v, d=0
if y = 0: n = 0, d=v
==>
n + d - v = 0
-G * (y(i)) <= n <= G * (y(i)) : objective coefficient
-G * (1-y(i))<= d <= G * (1-y(i)) : slack variable
==>
n + d - v = 0
0 <= n + G * y(i)
n - G * y(i) <= 0
0 <= d + (1- y(y(i)) * G
d - (1 - y(i)) * G <= 0
it is important to only create controls for coefficients that exist in the library
'''
#! This is a new addition and may change the way RD works. Added to fix target / lib comparison failure.
names = self._confirmControlTargets(lib, names)
limits = model.getColumnLimits()
rName = prefix + "reg__"
jmodel = LinearModel()
equation = "%s:n + d + %sI= 0" % (prefix+ "reg__%s",dir)
jmodel = self.parseEquations.applyRelation(jmodel,rName,equation,names,prefixPattern=prefix+"%s__")
imodel = LinearModel()
rName = prefix + "reg__"
for name in names:
if name in limits:
limit = limits[name]
else:
limit = (self.negfluxControlBound,self.posfluxControlBound)
vName = name
rName = prefix + "reg__" + name
cName = prefix + "n__" + name
dName = prefix + "d__" + name
imodel.addColumnLimit(cName,(None,None))
imodel.addColumnLimit(dName,(None,None))
#!imodel.addColumnLimit(cName,(limit[0],limit[1]))
#!imodel.addColumnLimit(dName,(limit[0],limit[1]))
imodel.addRowLimit(rName,(0.0,0.0))
imodel.addData(rName, vName, dir)
imodel.addData(rName, cName, 1.0)
imodel.addData(rName, dName, 1.0)
if not jmodel == imodel:
print "----------------------------apply equation failure-----------------------------"
model.addConstraints(imodel)
'''
Z^hybrid = C^nat*v + C^adj*n
'''
objective = {}
originalObjective = model.getObjective()
objective.update(originalObjective)
imodel = LinearModel()
#Set coefficients for controls
for name in names:
iname = prefix + "n__" + name
coeffecent = self._getRegulationCoeffecent(lib,name)
iCoeffecent = float(coeffecent) * float(self.regCoeffecent)
objective[iname] = -iCoeffecent
imodel.addData("adjValue",iname,-iCoeffecent)
#This constraint maybe cause of failure because of overloading of coefficients * fluxes in dual
#This constraint combined with the use of the control penalty causes infeasibility issues
#imodel.addRowLimit("adjValue",(None,None))
#model.addConstraints(imodel)
jmodel = LinearModel()
equation = "adjValue%s:objectiveVar + adjVar = 0"
jmodel = self.parseEquations.applyRelation(jmodel,relationName="adjValue",value=equation,targets=[''],prefixPattern="%s")
del jmodel.columnLimits["objectiveVar"]
jmodel.addColumnLimit("objectiveVar",(None,None))
imodel = LinearModel()
imodel.addData("adjValue","objectiveVar",1)
imodel.addData("adjValue","adjVar",1)
imodel.addRowLimit("adjValue",(0,0))
imodel.addColumnLimit("adjVar",(None,None))
imodel.addColumnLimit("objectiveVar",(None,None))
if not jmodel == imodel:
print "---------------------------------apply equation faliure: control objective-----------------------------------------"
#This constraint maybe cause of failure because of overloading of coefficients * fluxes in dual
#This constraint along with a control penalty causes infeasibility issues in the optimization
#model.addConstraints(imodel)#! consider removing to insure that adjValue is not messing up the system
model.setObjective(objective)
return model
def _constructDualRegulation(self,model,names,prefix):
'''
Construct limit / freedom variable for dual
These are the dual analog control variables
S'lambda + ... - delta - c' = 0
'''
rName = self.dualRName + prefix
jmodel = LinearModel()
equation = "%s" % (self.cdName)
jmodel = self.parseEquations.applyRelation(jmodel,rName,equation,names,prefixPattern="%s"+prefix)
imodel = LinearModel()
for name in names:
rName = self.dualRName + prefix + name
sName = self.cdName + prefix + name
imodel.addData(rName,sName,1)
imodel.addColumnLimit(sName,(None,None))
if not imodel == jmodel:
print "---------------------------------apply equation faliure: Dual control-----------------------------------------"
model.addConstraints(imodel)
return model
def _constructControl(self,model,controlNames,binaryPrefix,controlValuePref,regPrefix,climits,zero):
'''
Create binary control variables
one = off:
if y = 0: low < v < high
if y = 1: v = 0
v + high*y <= high
v + low*y >= low
zero = off:
if y = 1: 0 < v < d
if y = 0: v = 0
v - high*y <= 0
v - low*y >= 0
==>
for v in reaction targets
pc: v - high*pc < 0
nc: v - low* < 0
'''
binaryLimitName = "control_"
posControlRow = regPrefix+"_pc_"
negControlRow = regPrefix+"_nc_"
jmodel = LinearModel()
if zero:
equation1 = "%s_pc_%s:%s + %s %s < 0" % (regPrefix,"%s",controlValuePref,-1.0*self.posControlBound,binaryPrefix)
equation2 = "%s_nc_%s:%s + %s %s > 0" % (regPrefix,"%s",controlValuePref,-1.0*self.negControlBound,binaryPrefix)
else:
equation1 = "%s_pc_%s:%s + %s %s < %s" % (regPrefix,"%s",controlValuePref,self.posControlBound,binaryPrefix,self.posControlBound)
equation2 = "%s_nc_%s:%s + %s %s > %s" % (regPrefix,"%s",controlValuePref,self.negControlBound,binaryPrefix,self.negControlBound)
jmodel = self.parseEquations.applyRelation(jmodel,posControlRow,equation1,controlNames,prefixPattern="%s")
jmodel = self.parseEquations.applyRelation(jmodel,negControlRow,equation2,controlNames,prefixPattern="%s")
result = LinearModel()
for columnName in controlNames:
iCName = binaryPrefix + columnName
iDName = controlValuePref + columnName
posControlName = regPrefix + "_pc_" + columnName
negControlName = regPrefix + "_nc_" + columnName
if columnName in climits and False: # this method is currently broken
(negBound,posBound) = climits[columnName]
if negBound == None:
negBound = self.negControlBound
if posBound == None:
posBound = self.posControlBound
else:
posBound = self.posControlBound
negBound = self.negControlBound
'''
y = 1: l < v < u
'''
if(zero):
result.addData(posControlName,iDName,1.0)
result.addData(negControlName,iDName,1.0)
result.addData(posControlName,iCName,-posBound)
result.addData(negControlName,iCName,-negBound)
result.addRowLimit(posControlName,(None,0.0))
result.addRowLimit(negControlName,(0.0,None))
else:
result.addData(posControlName, iDName, 1.0)
result.addData(negControlName, iDName, 1.0)
result.addData(posControlName, iCName, posBound)
result.addData(negControlName, iCName, negBound)
result.addRowLimit(posControlName,(None,posBound))
result.addRowLimit(negControlName,(negBound,None))
jmodel.columnLimits = {}
if not result == jmodel:
#print "---------------------------------apply equation failure: control-----------------------------------------"
#! currently apply version does not handleA a map/list of boundary values
pass
model.addConstraints(result)
'''
w is limit of controls
y is actual control state variable
w - y = 0 as starting state ie all controls default to 0
sum w > control min
sum w < control max
'''
self.parseEquations.defaultLimit = (0.0,1.0)
iCName = binaryPrefix
iDName = controlValuePref
rowName = "_rcr_" + binaryPrefix
equation = "_rcr_%s:_rc_%s + -1.0 %s = 0" % (binaryPrefix+"%s",iCName,iCName)
jmodel = LinearModel()
jmodel = self.parseEquations.applyRelation(jmodel,rowName,equation,controlNames,prefixPattern="%s")
jmodel.setMipColumnNames(controlNames,tag=iCName+"%s")
self.parseEquations.defaultLimit = (None,None)
result = LinearModel()
for columnName in controlNames:
iCName = binaryPrefix + columnName
iDName = controlValuePref + columnName
result.setMipColumnName(iCName)
result.addColumnLimit(iCName,(0.0,1.0))
result.addColumnLimit("_rc_" + iCName,(0.0,1.0))
result.addData("_rcr_" + iCName, "_rc_" + iCName, 1.0)
result.addData("_rcr_" + iCName, iCName, -1.0)
result.addRowLimit("_rcr_" + iCName, (0.0,0.0))
if not result == (jmodel):
print "---------------------------------apply equation failure: control-----------------------------------------"
for columnName in controlNames:
iCName = binaryPrefix + columnName
iDName = controlValuePref + columnName
result.addData(binaryLimitName,"_rc_"+ iCName,1.0)
#Limit on control variables
result.addRowLimit(binaryLimitName,(self.controlMin,self.controlMax))
model.addConstraints(result)
return model
def _constructPDControlRelation(self,model,controlNames,controlPrefix,variablePrefix,rowPrefix,zeroed,controlLibraryName):
'''
Construct primal and dual control limitation relation to binary control variable.
'''
controlNames = self._confirmControlTargets(controlLibraryName, controlNames)
result = model
pRowPrefix = "__p_" + rowPrefix + "__" #prime control
dRowPrefix = "__d_" + rowPrefix + "__" #dual control
result = self._constructControl(result, controlNames, controlPrefix, variablePrefix, pRowPrefix, self.primeControlLimits, zeroed)
result = self._constructControl(result, controlNames, controlPrefix, "__sigma__" + variablePrefix, dRowPrefix, self.dualControlLimits, not zeroed)
result = self._constructDualRegulation(result,controlNames,variablePrefix)
return result
def _constructControlRestriction(self,model,names,prefix):
'''
Build restriction that one control library can be used for a control type.
Hopefully limit search space somewhat.
updated 8 12 08
0 < y + ny < 1
'''
scLimit = float(self.simoControl)
result = model
for name in names:
vname = prefix + name
rowName = "__rest_row__" + name
model.addData(rowName,vname, 1)
#model.addRowLimit(rowName,(0,1))
model.addRowLimit(rowName,(0,scLimit))
return result
def _constructControlPenalty(self,model,controlNames,prefix):
'''
This method can cause optimization infeasiblity problems
if variable boundaries are to large or model is large.
cp = y
objective = objective + delta*cp
'''
result = model
cpRow = "r_control_pentalty"
cpColumn = "c_control_penalty"
for name in controlNames:
iName = prefix+name
result.addData(cpRow,iName,1.0)
#These relations and controls are repeatedly created, so far does not cause problems
result.addData(cpRow,cpColumn,-1.0)
result.addRowLimit(cpRow,(0,0))
result.addColumnLimit(cpColumn,(0,None))
objective = model.getObjective()
iObjective = {}
iObjective.update(objective)
iObjective.update({cpColumn:self.controlDelta})
result.setObjective(iObjective)
return result
def _controlValueLimit(self,model,controlNames,postfix):
'''
Limits the absolute value of control effects
from any single flux control
Sum(t)Y(tj)*beta(tj) <= k
'''
result = model
if self.simoLimit == 0:
return model
for name in controlNames:
cpRow = "r_con_val_limit_%s" % name
for (libName,libPrefix) in self.libdata:
controlValue = self._getRegulationCoeffecent(libName, name)
iName = libPrefix+postfix+name
result.addData(cpRow,iName,controlValue)
cvLimit = (-self.simoLimit,self.simoLimit)
result.addRowLimit(cpRow,cvLimit)
return result
def constructControl(self,lp,controlNames, prefix, controlLibraryName):
result = lp
#! This is a new addition and may change the way RD works.
controlNames = self._confirmControlTargets(controlLibraryName, controlNames)
#Construct control relation in primal form of bi-level model
result = self._constructPDControlRelation(result, controlNames, prefix + "y__", prefix + "n__" , prefix + "p_", True, controlLibraryName)
#Construct control relation in dual form of bi-level model
result = self._constructPDControlRelation(result, controlNames, prefix + "y__", prefix + "d__" , prefix + "g_", False, controlLibraryName)
#Restriction that positive and negative part of control cannot be used together: !seems to be broken taken out for testing!
result = self._constructControlRestriction(result, controlNames, prefix + "y__",)
#Creates objective for minimizing controls, causes infeasible solution in larger models
if self.useControlPenalty:
result = self._constructControlPenalty(result,controlNames, prefix + "y__")
return result
def _getGeneNames(self):
return self.geneList
def _getReactionGenePairs(self):
result = []
for (rxn,genes) in self.rGeneMap.items():
for gene in genes:
result.append((rxn,gene))
return result
def geneControl(self,lp,prefix,controlNames,libName):
'''
Gene level control optimization relations
geneData in format [(rxn name, gene name)]
for reaction i in mapping to gene g:
reaction control(i) - gene control(g) = 0
or
rc(i) - gc(g) > 0
rc(i) - gc(g) < 0
'''
geneData = self._getReactionGenePairs()
libData = self.regulationLibrary[libName]
controlNames = self._confirmControlTargets(libName, controlNames)
binaryLimitName = "control_"
geneBinaryLimitName = "geneControl_"
geneControlNames = set()
reactionControlNames = set()
if geneData == None:
return lp
iGeneData = []
for (key,value) in geneData:
if key in controlNames:
iGeneData.append((key,value))
geneData = iGeneData
for (key,value) in geneData:
if key not in controlNames:
continue
if value == None:
print "Failed to map (%s, %s)" % (key,value)
continue
if key not in libData.keys():
if self.verbose: print "In control Library [%s] gene [%s] linked to rxn [%s] no control value" % (libName,value,key)
continue
equalRowName = "_meta_gene_" + prefix + key
lowRowName = "_low_meta_gene_" + prefix + key
highRowName = "_high_meta_gene_" + prefix + key
reactionControl = prefix + key
geneControl = prefix + value
geneControlNames.add(geneControl)
reactionControlNames.add(reactionControl)
#-----------------------------------------------------------------------------
# Scip previously could not cannot handle this binary equality
# This may have been a large variable problem and could currently be solved.
#-----------------------------------------------------------------------------
if self.unityControl:
lp.addData(equalRowName,reactionControl,1.0)
lp.addData(equalRowName,geneControl,-1.0)
lp.addRowLimit(equalRowName,(0,0))
else:
lp.addData(lowRowName,reactionControl,1.0)
lp.addData(lowRowName,geneControl,-1.0)
lp.addRowLimit(lowRowName,(0,None))
lp.addData(highRowName,reactionControl,1.0)
lp.addData(highRowName,geneControl,-1.0)
lp.addRowLimit(highRowName,(None,0))
equation1 = "%s: %s + -1.0 %s > 0" % (lowRowName,reactionControl,geneControl)
equation2 = "%s: %s + -1.0 %s < 0" % (highRowName,reactionControl,geneControl)
'''
w is limit of change in controls
y is actual control state variable
sum w > control min
sum w < control max
'''
for name in geneControlNames:
lp.addColumnLimit(name,(0,1))
lp.addColumnLimit("_rc_" + name,(0.0,1.0))
lp.setMipColumnName(name)
lp.setMipColumnName("_rc_" + name)
lp.addData("_rcr_" + name, "_rc_" + name, 1.0)
lp.addData("_rcr_" + name, name, -1.0)
lp.addRowLimit("_rcr_" + name, (0.0,0.0))
lp.addData(geneBinaryLimitName,"_rc_"+ name,1.0)
#Limit on control variables
lp.addRowLimit(binaryLimitName,(None,None))
lp.addRowLimit(geneBinaryLimitName,(self.controlMin,self.controlMax))
return lp
def construct(self,model,controlNames,newObjective):
'''
Top Level Function
Build Complete Framework
'''
self.newObjective = newObjective
result = LinearModel()
result.addModel(model)
sObjective = self._scaleMap(model.getObjective(),self.objectiveCoeffecent)
sObjective[self.newObjective] = .01
result.setObjective(sObjective)
for (libName,libPrefix) in self.libdata:
result = self._constructRegulationObjecive(result, controlNames, libName, libPrefix, -1.0)
pass
result = self._primalDual(result, result.getObjective())
for (libName,libPrefix) in self.libdata:
result = self.constructControl(result,controlNames,libPrefix,libName)
if self.rGeneMap != None:
self.geneControl(result,libPrefix + "y__",controlNames,libName)
if self.simoLimit !=0:
result = self._controlValueLimit(result,controlNames,postfix="y__")
return result
def _aproxEqual(self,v1,v2,delta):
return abs(v1 - v2) < delta
def _updateIterationControl(self,model,names,prefix,prediction,controlLibrary):
'''
updates framework status to new neighborhood point.
w,y in {0,1}
y is active / inactive variable
w is search limitation variable
if y = 1 and w = 0: control is on and unchanged:
w + y = 1
if y = 0 and w = 0: control is off, return to state w=y:
w - y = 0
if y = 1 and w = 1: Means control has been turned on, change to default y = 1, when w = 0
w + y = 1
if y = 0 and w = 1: Means control has been turned off, change default to y = 0 and w = 0
w - y = 0
y = prefix + name
w = _rc_ + prefix + name
'''
for name in names:
iCName = prefix + name #Acitive control
tCName = "_rc_" + iCName #Change control
if iCName not in prediction.keys():
continue
else:
value = prediction[iCName]
if tCName not in prediction.keys():
if self.verbose: print "Analog slack variable %s not found" % (tCName)
else:
tValue = prediction[tCName]
value = round(value,4)
tValue = round(tValue,4)
#!if value == 1.0 and tValue == 1.0:
if value == 1.0:
#if self.verbose: print "Making %s default active" % (name)
model.addData("_rcr_" + iCName, tCName, 1.0)
model.addData("_rcr_" + iCName, iCName, 1.0)
model.addRowLimit("_rcr_" + iCName, (1.0,1.0))
#equation = "_rcr_%s:%s + %s = 1.0" % (iCName,tCName,iCName)
#!if value == 0.0 and tValue == 1.0:
if value == 0.0:
model.addData("_rcr_" + iCName, tCName, 1.0)
model.addData("_rcr_" + iCName, iCName, -1.0)
model.addRowLimit("_rcr_" + iCName, (0.0,0.0))
#equation = "_rcr_%s:%s + %s = 0.0" % (iCName,tCName,iCName)
return model
def setControl(self,model,prefix,targets,on,binary=False):
#! merge with iterative control update
'''
updates framework status to new neighborhood point.
w,y in {0,1}
if y = 1 and w = 1: Means control has been turned on, change to default y = 1, when w = 0
w + y = 1
if y = 0 and w = 1: Means control has been turned off, change default to y = 0 and w = 0
w - y = 0
y = prefix + name
w = _rc_ + prefix + name
'''
for name in targets:
iCName = prefix + name
tCName = "_rc_" + iCName
if iCName not in model.getColumnNames():
if self.verbose: print "%s control not found" % (iCName)
continue
if tCName not in model.getColumnNames():
if self.verbose: print "%s slack control not found" % (tCName)
continue
if on == True: #Turn control on
if binary:
model.addColumnLimit(iCName,(1.0,1.0))
model.addColumnLimit(tCName,(0.0,1.0))
else:
model.addColumnLimit(iCName,(0.0,1.0))
model.addColumnLimit(tCName,(0.0,1.0))
model.addData("_rcr_" + iCName, iCName, 1.0)
model.addData("_rcr_" + iCName, tCName, 1.0)
model.addRowLimit("_rcr_" + iCName, (1.0,1.0))
if on == False: # Turn control off
if binary:
model.addColumnLimit(iCName,(0.0,1.0))
model.addColumnLimit(tCName,(0.0,1.0))
else:
model.addColumnLimit(iCName,(0.0,1.0))
model.addColumnLimit(tCName,(0.0,1.0))
model.addData("_rcr_" + iCName, tCName, 1.0)
model.addData("_rcr_" + iCName, iCName, -1.0)
model.addRowLimit("_rcr_" + iCName, (0.0,0.0))
#equation = "_rcr_%s:%s - %s = 0.0" % (iCName,tCName,iCName)
return model
def activateControl(self,model,libControl,on=True):
for(libName,targets) in libControl.items():
prefix = self.controlPrefixMap[libName]
self.setControl(model, prefix+"y__", targets, on)
return model
def iterate(self,model,targets,libdata,prediction):
for (libName,prefix) in libdata:
if self.rGeneMap != None:
itargets = self.geneList
else:
itargets = self._confirmControlTargets(libName, targets)
self._updateIterationControl(model,itargets,prefix + "y__",prediction, libName)
return model
def findControl(self,controlMap):
result = []
for (key,prefix) in self.libdata:
valueMap = self.regulationLibrary[key]
ukeys = set(valueMap.keys()).intersection(controlMap.keys())
for uk in ukeys:
if valueMap[uk] == controlMap[uk]:
result.append((uk,prefix,valueMap[uk]))
return result
def generateControlObjective(self,activeControl):
yControlNames = []
wControlNames = []
controlLibrary = self.findControl(activeControl)
for (key,prefix,value) in controlLibrary:
if self.rGeneMap == None or len(self.rGeneMap) == 0:
yControlName = prefix + "y__" + key
yControlNames.append(yControlName)
wControlNames.append("_rc_" + yControlName)
elif self.rGeneMap != None:
if key in self.rGeneMap.keys():
gNames = self.rGeneMap[key]
for gName in gNames:
yControlName = prefix + "y__" + gName
yControlNames.append(yControlName)
wControlNames.append("_rc_" + yControlName)
result = {}
for i in yControlNames:
result[i] = 1.0
for i in wControlNames:
result[i] = 1.0
return result
def iterateObjectiveControl(self,model,targets,prefix,libdata,objective,prediction):
#controls = self._activateControl(prefix,objective.keys())
model = self.iterate(model,targets,libdata,prediction)
return model
def generateCoeffecents(self,targets,prefix,predictions,controlName):
result = {}
for name in targets:
iCName = prefix + name
if iCName not in predictions:
continue
value = predictions[iCName]
if value >= 1.0 - self.delta:
controlValue = self._getRegulationCoeffecent(controlName, name)
controlValue = -1.0 * controlValue
result[name] = controlValue
return result
def _convertToGeneNames(self,reactionName):
if reactionName not in self.rGeneMap.keys():
return None
geneNames = self.rGeneMap[reactionName]
if type(geneNames) == type(""):
return set([geneNames])
else:
return geneNames
def _convertToGeneTag(self,reactionName,geneClusters = None):
geneNames = self._convertToGeneNames(reactionName)
if geneNames == None:
return reactionName
iGeneNames = set(geneNames)
for geneName in geneNames:
if geneClusters != None:
if geneName in geneClusters.keys():
for iGeneName in geneClusters[geneName]:
iGeneNames.add(iGeneName)
geneTag = ''
for iGeneName in iGeneNames:
geneTag = geneTag + " " + iGeneName
geneTag = geneTag[1:]
return geneTag
def printGeneObjective(self,iObjective,geneClusters=None):
iGeneObjective = {}
iOtherObjective = {}
igControl = {}
for rxnName in iObjective.keys():
rControl = iObjective[rxnName]
if rxnName not in self.rGeneMap.keys():
iOtherObjective[rxnName] = rControl
else:
geneTag = self._convertToGeneTag(rxnName, geneClusters)
if geneTag not in iGeneObjective:
igControl[geneTag] = []
igControl[geneTag].append(rControl)
for k in igControl.keys():
iGeneObjective[k] = mean(igControl[k])
return (iGeneObjective,iOtherObjective)
def generateObjective(self,model,targets,predictions):
result = {}
libresult = {}
objective = model.getObjective()
for name in objective.keys():
value = objective[name]
result[name] = value * self.objectiveCoeffecent
for pair in self.libdata:
libName = pair[0]
prefix = pair[1]
controls = self.generateCoeffecents(targets,prefix + "y__",predictions,libName)
libcontrols = {}
for key in controls.keys():
libcontrols[(libName,key)] = controls[key]
if key not in result.keys():
result[key] = 0
ir = result[key]
result[key] = ir + controls[key]
libresult.update(libcontrols)
return (result,libresult)
def getControlNames(self,targets,predictions,prefix=True):
rControls = []
for pair in self.libdata:
libName = pair[0]
libPrefix = pair[1]
iControls = self._getNonZeroRegulationNames(libName,targets)
controls = self.generateCoeffecents(targets,libPrefix + "y__",predictions,libName)
iControls = controls
if self.rGeneMap != None:
zControls = set()
for controlName in iControls:
if controlName in self.rGeneMap.keys():
for gcn in self.rGeneMap[controlName]:
zControls.add(gcn)
iControls = zControls
for controlName in iControls:
if prefix:
rControl = (libName,libPrefix,controlName)
else:
rControl = (libName,controlName)
if rControl not in rControls:
rControls.append(rControl)
controlNames = rControls
#tcontrolNames = controlNames.intersection(self.targets)
return controlNames
def _createReportItem(self,predictions,names,prefix):
'''
Create report item from framework result
'''
result = {}
for name in names:
tName = prefix + name
if tName in predictions:
result[name] = predictions[tName]
return result
def getFluxValues(self,predictions,fluxNames):
return self._createReportItem(predictions, fluxNames, self.pName)
def createReport(self, predictions, fluxNames, targets, geneReduction = None):
'''
Create report from framework result
'''
fluxNames.append("objectiveVar")
fluxNames.append("objectiveVarFull")
report = Report()
report["flux"] = self._createReportItem(predictions, fluxNames, self.pName)
geneSet = set()
if self.rGeneMap != None:
for (key,values) in self.rGeneMap.items():
geneSet = geneSet.union(set(values))
for (libName,prefix) in self.libdata:
controlMap = self.regulationLibrary[libName]
report[libName] = self._createReportItem(predictions,targets,prefix+"y__")
report[libName+"_Value"] = controlMap
if self.rGeneMap != None:
geneControl = self._createReportItem(predictions,geneSet,prefix+"y__")
if geneReduction != None:
reducedGeneValues = {}
for k in geneControl.keys():
if k in geneReduction.keys():
for reducedGene in geneReduction[k]:
reducedGeneValues[reducedGene] = geneControl[k]
geneControl.update(reducedGeneValues)
r1 = report[libName]
r1.update(geneControl)
report[libName] = r1
#report["BioControl"] = self._createReportItem(predictions,targets,"__y__")
#report["adjustment pos"] = self._createReportItem(predictions,targets,self.cvName)
#report["difference pos"] = self._createReportItem(predictions,targets,"__d__")
#report["BioCoeffecents"] = self.posCoeffecentMap
#report["SynControl"] = self._createReportItem(predictions,targets,"__ny__")
#report["adjustment neg"] = self._createReportItem(predictions,targets,"__nn__")
#report["difference neg"] = self._createReportItem(predictions,targets,"__nd__")
#report["SynCoeffecents"] = self.negCoeffecentMap
return report