Package 'CellNOptR'

Description

This package does optimisation of boolean logic networks of signalling pathways based on a previous knowledge network and a set of data collected upon perturbation of some of the nodes in the network.

Details

Author(s)

T. Cokelaer, A. MacNamara, F. Eduati, S. Schrier, C. Terfve

Maintainer: A. Gabor<[email protected]>, until 2018-01-18: T. Cokelaer <[email protected]>

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

Examples

# quick 1 time point optimisation of a Prior Knowledge Network to MIDAS data. 
data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

pknmodel = ToyModel
cnolist = CNOlist(CNOlistToy)
model = preprocessing(cnolist, pknmodel)
results = gaBinaryT1(cnolist, model, verbose=FALSE)
plotFit(results)
cutAndPlot(cnolist, model, list(results$bString))

# Same as above and HTML report
CNORwrap(name="Toy", 
    namesData=list(CNOlist="ToyData",model="ToyModel"),
    data=CNOlistToy, model=pknmodel)

Description

Build a SIF table from a logic rule written in a string

Usage

build_sif_table_from_rule(rule_str, target, last_and_num = 0)

Arguments

Value

data.frame with the network structure derived from the rule. The column 'sif_str' contains the string that can be written to a file and then read with 'readSIF()' in order to load a CellNOpt compatible network.

Examples

CellNOptR:::build_sif_table_from_rule("B & (C | D)", "A", last_and_num=2)

test_rule <- list()
test_rule[[1]] <- "AMP_ATP | (ATM & ATR) | HIF1 | !(EGFR | FGFR3)"
test_rule[[2]] <- "A & ((B | C) & !(D & E))"
test_rule[[3]] <- "A & B | C"
test_rule[[4]] <- "A & B & C"
test_rule[[5]] <- "A & (B | C)"
test_rule[[6]] <- "(A | B) & (C | D)"
test_rule[[7]] <- "!(C & D) | (E & F)"
test_rule[[8]] <- "(A | B) & (C | !D) & (E | F)"
parsed_rule <- list()
for (i in c(1:length(test_rule))){
  parsed_rule[[i]] <- CellNOptR:::build_sif_table_from_rule(test_rule[[i]], "T")
}

Description

This function checks that all the signals in a CNOlist match to species in the model. It also checks that the CNOlist and Model lists have the right format and contain the right fields. It is called by the preprocessing function so there is no need to call it directly anymore if you use the preprocessing function.

In version 1.3.20, check of inhibitors and stimuli is also performed.

Usage

checkSignals(CNOlist, model)

Arguments

Details

If the formats are wrong, this function produces an error. If the signals/inhibitors/stimuli do not match the species, this function produces a warning that explains which signals does no match any species.

Author(s)

C. Terfve, T. Cokelaer

See Also

makeCNOlist, readSIF, preprocessing

Examples

data(CNOlistToy, package="CellNOptR")
data(ToyModel, package="CellNOptR")
checkSignals(CNOlistToy, ToyModel)

Description

This function fetch a file from the URL provided (default is http://www.ebi.ac.uk/~cokelaer/cellnopt/data/_downloads) and save it into a temporary file.

You will need Rcurl package to use this function.

Usage

CNOdata(filename, verbose=FALSE, url=NULL)

Arguments

Value

the path of the temporary file where data has been copied.

Author(s)

T. Cokelaer

Examples

## Not run: readSIF(CNOdata("PKN-ToyMMB.sif"))

Description

This function takes as input the filename of a MIDAS file (or the list returned by makeCNOlist) and returns an instance of CNOlist class. It provides an object oriented approach to manipulate CNOlist. This function calls readMIDAS and makeCNOlist.

Objects from the Class

Objects can be created by calls of the form new("CNOlist", ...).

Slots

cues:

Object of class "matrix"

inhibitors:

Object of class "matrix"

stimuli:

Object of class "matrix"

signals:

Object of class "list"

variances:

Object of class "list"

timepoints:

Object of class "vector" timepoints contained in the signals matrix.

See CNOlist-methods for details

Methods

Available methods are plot, compatCNOlist, randomize, length. See CNOlist-methods for details.

Author(s)

T. Cokelaer

See Also

randomizeCNOlist, plotCNOlist, plotCNOlist2

Examples

showClass("CNOlist")

files<-dir(system.file("ToyModel",package="CellNOptR"),full=TRUE)
cnolist  = CNOlist(files[[1]])
# getters:
getCues(cnolist)
getInhibitors(cnolist)
getSignals(cnolist)
getVariances(cnolist)
getTimepoints(cnolist)
getStimuli(cnolist)
# In version 1.3.30 and above, use the plot method instead of former plotCNOlist function.
plot(cnolist)
new_cnolist = randomize(cnolist)
length(cnolist)

List of CNOlist-class methods

Description

CNOlist is a class with a set of methods described here below.

Usage

signature(x="CNOlist")

Getters

getCues

Returns the cues (matrix) found in the CNOlist

getSignals

Returns the signals (list of matrices) found in the CNOlist

getStimuli

Returns the cues found in the CNOlist

getInhibitors

Returns the inhibitors found in the CNOlist

getTimepoints

Returns the timepoints found in the CNOlist

getVariances

Returns the Variances (list of matrices) found in the CNOlist. Will be different from zero only if replicates were found in the MIDAS data. See makeCNOlist

Setters

setSignals

Set the signals. No sanity check!

Other methods

compatCNOlist

convert the instance CNOlist into the old-style returned by makeCNOlist that is a list. Used in the ODE package.

length

returns length of CNOlist (number of time points)

randomize

randomizes the signals matrice in a CNOlist. See randomizeCNOlist for details

show

prints summary information

plot

plot the CNOlist instance using the plotCNOlist function.

plot

signature(x="CNOlist", y="CNOlist"): Please see the page of plotCNOlist2 for more details.

readErrors

signature(object="CNOlist", filename="MIDAS-file"): reads measurement error corresponding to the data from the MIDAS file and updates the CNOlist object

writeErrors

signature(object="CNOlist", filename="string",overwrite=F): writes measurement error corresponding to the data from the CNOlist to a MIDAS file

Author(s)

T.Cokelaer

See Also

CNOlist-class, randomizeCNOlist makeCNOlist

Examples

showClass("CNOlist")

data(CNOlistToyMMB, package="CellNOptR")
cnolist = CNOlistToyMMB

# In version 1.3.30 and above, use the plot method instead of former plotCNOlist function.
plot(cnolist)
# In version 1.5.14 and above, use getters instead of the @ operator
getCues(cnolist)

# others:
new_cnolist = randomize(cnolist)

Description

This data object contains the DREAM data used in the package vignette, already loaded and formatted as a CNOlist object. This is to be used with the model "DreamModel". This is a data collected on HepG2 cells cultivated with or without stimulation of tgfa, ilk, mek12, pi3k and p38, in combination with inhibition of igf1 and/or il1a. Seven phosphoproteins are measured using Luminex xMAP assays: akt, erk12, ikb, jnk12, p38, hsp27 and mek12.

Usage

data(CNOlistDREAM)

Format

CNOlistDREAM is a list with the fields "namesCues" (character vector), "namesStimuli" (character vector), "namesInhibitors" (character vector), "namesSignals" (character vector), "timeSignals" (numerical vector), "valueCues" (numerical matrix), "valueInhibitors" (numerical matrix), "valueStimuli"(numerical matrix), "valueSignals"(numerical matrix).

Source

This data and model is extracted from the Matlab version of CellNOpt1.0 (http://www.ebi.ac.uk/saezrodriguez/software.html#CellNetOptimizer).

References

1. J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

2. Prill RJ, Marbach D, Saez-Rodriguez J, Sorger PK, Alexopoulos LG, Xue X, Clarke ND, Altan-Bonnet G, and Stolovitzky G. Towards a rigorous assessment of systems biology models: the DREAM3 challenges. PLoS One, 5(2):e9202, 2010.

Description

This data object contains the data associated with the Toy Model example from the package vignette, already loaded and formatted as a CNOlist object.

Usage

data(CNOlistToy)

Format

CNOlistToy is a list with the fields "namesCues" (character vector), "namesStimuli" (character vector), "namesInhibitors" (character vector), "namesSignals" (character vector), "timeSignals" (numerical vector), "valueCues" (numerical matrix), "valueInhibitors" (numerical matrix), "valueStimuli"(numerical matrix), "valueSignals"(numerical matrix).

Source

This data and model is extracted from the Matlab version of CellNOpt1.0 (http://www.ebi.ac.uk/saezrodriguez/software.html#CellNetOptimizer).

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

Description

This data object contains the data associated with the Toy Model example from the package vignette, already loaded and formatted as a CNOlist object, and modified to contain 2 time points. The second time point is such a way that all of the signals stay as in time 1, except for cJun and Jnk which go to zero.

Usage

data(CNOlistToy)

Format

CNOlistToy is a list with the fields "namesCues" (character vector), "namesStimuli" (character vector), "namesInhibitors" (character vector), "namesSignals" (character vector), "timeSignals" (numerical vector), "valueCues" (numerical matrix), "valueInhibitors" (numerical matrix), "valueStimuli"(numerical matrix), "valueSignals"(numerical matrix).

Source

This data and model is extracted from the Matlab version of CellNOpt1.0 (http://www.ebi.ac.uk/saezrodriguez/software.html#CellNetOptimizer).

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

Description

This function performs the optimisation of a PKN model to a CNOlist data set. It optimises each time point found in the data and returns the processed model as well as a list of optimised bitstring corresponding to each time points that has been optimised.

This function does not create any plots or reports unlike CNORwrap.

Usage

CNORbool(CNOlist, model, paramsList=defaultParameters(),
    compression=TRUE, expansion=TRUE, cutNONC=TRUE, verbose=FALSE,
    timeIndices = NULL)

Arguments

Value

This function returns 2 components. The first one is the processed model used in the optimisation. The second is a list of optimised bitstrings found for each time points available in the MIDAS data set.

Author(s)

T.Cokelaer, S.Schrier

Examples

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")
res = CNORbool(CNOlist=CNOlistToy, model=ToyModel)

Description

This function is a wrapper around the whole CNOR analysis, it performs the following steps:

1. Plot the CNOlist;

2. Checks data to model compatibility;

3. Call the preprocessing function (Cut the nonc off the model, compress the model and expand the gates);

4. Compute the residual error;

5. Prepare for simulation;

6. Optimisation T1 and T2 (optional);

7. Plot simulated and experimental results;

8. Plot the evolution of fit;

9. Write the scaffold and PKN;

10. Write the report

Usage

CNORwrap(paramsList=NA, data=NA, model=NA, name, namesData=NA, time=1,
compression=TRUE, expansion=TRUE, cutNONC=TRUE)

Arguments

and the Data and Model structure.

If paramsList is not provided (NA), it is filled internally with the defaultParameters function.

If Data and Model are not provided in paramList, the function looks for Data and Model arguments.

If Data and Model are provided, the function overwrites the field data and model in paramsList.

Details

If you do not provide a parameters list, you can provide only essential elements, and all other parameters will be set to their default values. In this case, you should set paramsList=NA, and provide the following fields: data, model, name, time.

Value

This function does not return anything, it does the analysis, produces all the plots and puts them in a folder that is in your working directory, and is called "Name".

Author(s)

C. Terfve

Examples

#version with paramslist

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

pList = defaultParameters(CNOlistToy, ToyModel)
pList$maxGens = 5
pList$popSize = 5

CNORwrap(paramsList=pList, name="Toy",
    namesData=list(CNOlist="ToyData", model="ToyModel"))

## Not run: 
#version with default parameters

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

CNORwrap(name="Toy",
    namesData=list(CNOlist="ToyData",model="ToyModel"),
	data=CNOlistToy, model=ToyModel)

## End(Not run)

Description

This function compresses a model by compressing species that are not signals/inhibited/stimulated and that are not dead ends/in complex logic (i.e. only species with either one input or one output are compressed)/in self loops.

You can also use preprocessing function instead that calls compressModel and other preprocessing functions.

Usage

compressModel(model, indexes)

Arguments

Details

Be aware that in the multiple inputs/one output case, if one of the outputs is an '&' gate this function handles it fine as long as it is an '&' with 2 inputs and no more.

Value

a compressed model list, with an additional field called 'speciesCompressed' that contains the names of the species that have been compressed

Note

No need to call this function directly since version 0.99.24. Use preprocessing instead.

Author(s)

C. Terfve

See Also

indexFinder, readSIF, preprocessing

Examples

#load data

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=FALSE)
toyComp<-compressModel(ToyModel,indicesToy)

Description

The bitString made of 0 and 1 allows to select a submodel from the model provided. Then, the simulator function are called to compute the objective function. The sizeFac and NAFac are penalties added to the final score as described in gaBinaryT1. The indexList and simList arguments can be provided to speed up the code otherwise, they are recomputed from the CNOlist and model.

Usage

computeScoreT1(CNOlist, model, bString, simList=NULL, indexList=NULL,
    sizeFac=0.0001, NAFac=1, timeIndex=2)

Arguments

Value

Author(s)

T. Cokelaer

Examples

data(CNOlistToy,package="CellNOptR")
    data(ToyModel,package="CellNOptR")
    model <- preprocessing(CNOlistToy,ToyModel)
    score = computeScoreT1(CNOlist(CNOlistToy), model, bString=rep(1,16))

Description

The bitString made of 0 and 1 allows to select a submodel from the model provided. Then, the simulator function are called to compute the objective function. The sizeFac and NAFac are penalties added to the final score as described in gaBinaryTN.

Usage

computeScoreTN(CNOlist, model, simList=NULL, indexList=NULL, simResPrev=NULL,
    bStringPrev=NULL, bStringNext=NULL, timeIndex=NULL, sizeFac=0.0001, NAFac=1, bStrings=NULL)

Arguments

Value

Author(s)

T. Cokelaer, S.Schrier

Examples

data(CNOlistToy2,package="CellNOptR")
    data(ToyModel2,package="CellNOptR")
    model <- preprocessing(CNOlistToy2, ToyModel2)
    bStringT1 = c(0,0,1,1,1,1,1,1,0,0,1,1,1,1,1,1)
    simT1<-simulateTN(CNOlist=CNOlistToy2, model=model, bStrings=list(bStringT1))

    score1 = computeScoreTN(CNOlistToy2, model, bStrings=list(bStringT1,c(1,0,1,0)))

Description

This function takes as input the model and data in a CNOlist object in order to define the set of binary variablese.

Usage

create_binaries(model, 
                            midas, 
                            numberOfExperiments, 
                            y_vector)

Arguments

Author(s)

E Gjerga, H Koch

References

Alexander Mitsos, Ioannis N. Melas, Paraskeuas Siminelakis, Aikaterini D. Chairakaki, Julio Saez-Rodriguez, and Leonidas G. Alexopoulos. Identifying Drug Effects via Pathway Alterations using an Integer Linear Programming Optimization Formulation on Phosphoproteomic Data. PLoS Comput Biol. 2009 Dec; 5(12): e1000591.

Description

This function takes as an input the cno inputs (model + data) together with CPLEX parameters and then solves the ILP problem.

Usage

createAndRunILP(model = model, 
                            midas = midas, 
                            cnolist = cnolist, 
                            accountForModelSize = accountForModelSize, 
                            sizeFac = sizeFac, 
                            source_path = source_path, 
                            mipGap = mipGap, 
                            relGap = relGap, 
                            timelimit = timelimit,
                            cplexPath = cplexPath, 
                            method = method, 
                            numSolutions = numSolutions, 
                            limitPop = limitPop, 
                            poolIntensity = poolIntensity, 
                            poolReplace = poolReplace)

Arguments

Author(s)

E Gjerga, H Koch

Description

This function takes as the cplex optimization results and the variables assigned to each interaction in the ILP formulation in order to read the optimal bitstring.

Usage

createILPBitstringAll(cplexSolutionFileName,
                                 y_vector,
                                 binary_variables)

Arguments

Author(s)

E Gjerga, H Koch

Description

If an inhibitor is also a measured species, replace the data with NA (when inhibited)

Usage

crossInhibitedData(object)

Arguments

Value

the new cnolist

Author(s)

T. Cokelaer

Examples

data(ToyModel,package="CellNOptR")
data(CNOlistToy,package="CellNOptR")
cnolist = crossInhibitedData(CNOlist(CNOlistToy))

Description

Cross-validation analysis for the boolean case.

Usage

crossvalidateBoolean(CNOlist,model,nfolds=10,foldid=NULL, 
                                type=c('datapoint','experiment','observable'),timeIndex = 2,parallel=FALSE,...)

Arguments

Details

Does a k-fold cross-validation for Boolean CellNOpt models. In k-iterations a fraction of the data is eliminated from the CNOlist. The model is trained on the remaining data and then the model predicts the held-out data. Then the prediction accuracy is reported for each iteration.

Value

This function returns a list with elements:

Author(s)

A. Gabor, E. Gjerga

Examples

data("ToyModel", package="CellNOptR")
data("CNOlistToy", package="CellNOptR")
pknmodel = ToyModel
cnodata = CNOlist(CNOlistToy)

# original and preprocessed network 
plotModel(pknmodel,cnodata)
model = preprocessing(data = cnodata,
					model = pknmodel,
					compression = TRUE,
					expansion = TRUE)
plotModel(model,cnodata)

# original CNOlist contains many timepoints, we use only a subset
plot(cnodata)
selectedTime = c(0,10)
cnodata_prep = cutCNOlist(cnodata,
                          model = model,
                          cutTimeIndices = which(!getTimepoints(cnodata) %in% selectedTime))

plot(cnodata_prep)

# optimise and show results
opt = gaBinaryT1(CNOlist = cnodata_prep,model = model,verbose = FALSE)

# 10-fold crossvalidation using T1 data
# We use only T1 data for crossvalidation, because data in the T0 matrix is not independent.
# All rows of data in T0 describes the basal condition.

# Crossvalidation produce some text in the command window:  
## Not run: 
library(doParallel)
registerDoParallel(cores=3)
R=crossvalidateBoolean(CNOlist = cnodata_prep,
						model = model,
						type = "datapoint",
						nfolds = 10,
						parallel = TRUE)


## End(Not run)

Description

This function takes a model and cnolist as well as a list of optimised bitstring at different time points. It calls the appropriate cutAndPlotResultsTX function.

Usage

cutAndPlot(CNOlist, model,  bStrings, plotPDF=FALSE, tag=NULL,
plotParams = list(maxrow = 10))

Arguments

Value

This function returns nothing. It plots a graph in your graphic window and sqve it in a file if asked

Author(s)

T. Cokelaer

See Also

cutAndPlotResultsT1

Examples

#load data

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

#pre-process model
model = preprocessing(CNOlistToy, ToyModel)

cutAndPlot(CNOlistToy, model,
    bStrings=list(rep(1,length(model$reacID))),	plotPDF=FALSE)

Description

This function takes a model and an optimised bitstring, it cuts the model according to the bitstring and plots the results of the simulation along with the experimental data.

Usage

cutAndPlotResultsT1(model, bString, simList=NULL, CNOlist, indexList=NULL, plotPDF =
FALSE, tag = NULL, tPt=CNOlist@timepoints[2],  plotParams = list(maxrow=10))

Arguments

Value

This function returns plotted MSEs and list of filenames generated (if any)

Author(s)

C.Terfve, T. Cokelaer,A. MacNamara

See Also

gaBinaryT1

Examples

#load data

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

#pre-process model
model = preprocessing(CNOlistToy, ToyModel)

#optimise
ToyT1opt<-gaBinaryT1(
	CNOlist=CNOlistToy,
	model=model,
	maxGens=20, 
	popSize = 10,
	verbose=FALSE)

#plotting
cutAndPlotResultsT1(
	model=model,
	CNOlist=CNOlistToy,
	bString=ToyT1opt$bString,
	plotPDF=FALSE)

Description

This function takes a model and an optimised bitstring, it cuts the model according to the bitstring and plots the results of the simulation along with the experimental data. This function is designed to work on results of a 2 step optimisation.

Usage

cutAndPlotResultsTN(CNOlist, model, bStrings, plotPDF = FALSE, tag=NULL, 
    plotParams = list(maxrow = 10))

Arguments

Value

This function returns plotted MSEs

Note

New in version 1.3.28

Author(s)

T. Cokelaer, A. MacNamara, Sarah Schrier, C. Terfve based on cutAndPlotResultsT1

See Also

gaBinaryT1, prep4sim, cutAndPlotResultsT1

Examples

#load data

data(CNOlistToy2,package="CellNOptR")
data(ToyModel2,package="CellNOptR")

#pre-process model
model = preprocessing(CNOlistToy2, ToyModel2)

#optimise t1
ToyT1<-gaBinaryT1(
	CNOlist=CNOlistToy2,
	model=model,
	maxGens=20, 
	popSize = 10,
	verbose=FALSE)

#Optimise T2
ToyT2<-gaBinaryTN(
	CNOlist=CNOlistToy2,
	model=model,
	bStrings=list(ToyT1$bString),
	maxGens=20, 
	popSize = 10,
	verbose=FALSE)

cutAndPlotResultsTN(
	CNOlist=CNOlistToy2,
	model=model,
	bStrings=list(ToyT1$bString, bStringT2=ToyT2$bString),
	plotPDF=FALSE)

Description

The MIDAS file may contain species that are not contained in the model. If you want to remove cues and signals from your CNOlist that are not contained in the model, you can use this function by using the parameter model. It can also be used to remove some time points by using the parameter cutTimeIndices. Both parameters can be used at the same time and at least one of them must be provided.

Usage

cutCNOlist(cnolist, model, cutTimeIndices, verbose=FALSE)

Arguments

Value

Note

added in version 1.5.10

Author(s)

T. Cokelaer

See Also

CNOlist-class

Examples

data(CNOlistToy,package="CellNOptR")
    data(ToyModel,package="CellNOptR")
    
    cno_prep = cutCNOlist(cnolist = CNOlistToy,model = ToyModel,verbose = FALSE)

Description

This function is for developers only.

Usage

cutModel(model, bString)

Arguments

Value

Note

added in version 1.3.16

Author(s)

T. Cokelaer

Examples

data(ToyModel,package="CellNOptR")
    
    bString = rep(1,length(ToyModel$reacID))
    # remove some reactions by setting them to 0. 
    bString[c(2,5,6)] <- 0
    
    prepModel = cutModel(model = ToyModel, bString = bString )

Description

This function cuts the non-observable and/or non-controllable species from the model, and returns a cut model.

Usage

cutNONC(model, NONCindexes)

Arguments

Details

This function takes in a model and a vector of indices of species to remove in that model and it removes those species and any reaction involving them (be aware, if you have x&y=z and x is to be removed, then the function produces y=z, because it works by removing entire rows of the model matrices and then removes the columns that do not have either an input or an output). This function could actually be used to cut any species, not only NONC species.

Value

a model

Note

No need to call this function directly since version 0.99.24. Use preprocessing instead.

Author(s)

C.Terfve

See Also

findNONC, readSIF

Examples

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")
indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=FALSE)
ToyNCNOindices<-findNONC(ToyModel,indicesToy,verbose=FALSE)
ToyNCNOcut<-cutNONC(ToyModel,ToyNCNOindices)

Description

This function is for developers only.

Usage

cutSimList(simList, bString)

Arguments

Value

Author(s)

T. Cokelaer

Description

This function provides a list of default parameters including the Genetic Algorithm parameters.

Usage

defaultParameters(data=NA, model=NA)

Arguments

Details

The list contains the Genetic Algorithm parameter, a verbose option and can be used to store the Data and Model.

Value

Author(s)

T. Cokelaer

Examples

data(ToyModel, package="CellNOptR")
data(CNOlistToy, package="CellNOptR")
params = defaultParameters(CNOlistToy, ToyModel)

Description

This function performs an exhaustive search of the parameter space tring all the solutions. It is used internally by the genetic algorithm when a small model has to be optimised and the number of solutions to try is smaller than the number of iterations that the Genetic Algorithm will perform.

Usage

exhaustive(CNOlist, model, shuffle=FALSE, Nmax=NULL, verbose=TRUE, sizeFac =
0.0001, NAFac = 1, relTol=0.1, timeIndex=2)

Arguments

Value

This function returns a list with elements:

Note that the field results, is not yet populated but maybe in the future.

Author(s)

T. Cokelaer

See Also

gaBinaryT1

Examples

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

#pre-process model

model = preprocessing(CNOlistToy, ToyModel)

#optimise

results <-exhaustive(
	CNOlist=CNOlistToy,
	model=model,
    shuffle=TRUE,
    Nmax=1000,
	verbose=FALSE)

Description

This function takes in a model and (1) splits all AND gates into ORs. In addition, (2) wherever there are more than one, it creates all possible ANDs combinations of them, but considering only ANDs with 2, 3 or 4 inputs according to the user argument (default is 2)

Usage

expandGates(model, ignoreList=NA, maxInputsPerGate=2)

Arguments

Details

This function returns a model with additional fields that help keep track of the processing done on the network. I would advice not to overwrite on the initial model but rather to assign the result of this function to a variable with a different name.

Value

returns a model, with additional fields:

Note

No need to call this function directly since version 0.99.24. Use preprocessing instead.

Author(s)

C.Terfve. T. Cokelaer, A.MacNamara, Martin-Franz-Xaver Pirkl

Examples

#load data

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

#pre-process the model

indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=TRUE)
ToyNCNOindices<-findNONC(ToyModel,indicesToy,verbose=TRUE)
ToyNCNOcut<-cutNONC(ToyModel,ToyNCNOindices)
indicesToyNCNOcut<-indexFinder(CNOlistToy,ToyNCNOcut)
ToyNCNOcutComp<-compressModel(ToyNCNOcut,indicesToyNCNOcut)
indicesToyNCNOcutComp<-indexFinder(CNOlistToy,ToyNCNOcutComp)
ToyNCNOcutCompExp<-expandGates(ToyNCNOcutComp, maxInputsPerGate=4)

Description

This function finds the indexes of the non-observable and non controllable species and returns the indices, in the model, of the species to remove

Usage

findNONC(model, indexes, verbose=FALSE)

Arguments

Details

This function uses the function floyd.warshall.all.pairs.sp from the package RBGL. Non observable nodes are those that do not have a path to any measured species in the model, whereas non controllable nodes are those that do not receive any information from a species that is perturbed in the data.

Value

a vector of indices of species to remove

Note

No need to call this function directly since version 0.99.24. Use preprocessing instead.

Author(s)

C. Terfve

See Also

cutNONC, indexFinder, readSIF

Examples

data(CNOlistToy, package="CellNOptR")
    data(ToyModel, package="CellNOptR")
    checkSignals(CNOlistToy, ToyModel)
    indicesToy <- indexFinder(CNOlistToy, ToyModel)
    ToyNCNOindices <- findNONC(ToyModel, indicesToy)

Description

This function is the genetic algorithm to be used to optimise a model by fitting to data containing one time point.

Usage

gaBinaryT1(CNOlist, model, initBstring=NULL, sizeFac = 1e-04, 
    NAFac = 1, popSize = 50, pMutation = 0.5, maxTime = 60, maxGens = 500,
    stallGenMax = 100, selPress = 1.2, elitism = 5, relTol = 0.1, verbose=TRUE,
    priorBitString=NULL, timeIndex=2)

Arguments

Details

The whole procedure is described in details in Saez-Rodriguez et al. (2009). The basic principle is that at each generation, the algorithm evaluates a population of models based on excluding or including some gates in the initial pre-processed model (this is encoded in a bitstring with contains 0/1 entries for each gate). The population is then evolved based on the results of the evaluation of these networks, where the evaluation is obtained by simulating the model (to steady state) under the various conditions present in the data, and then computing the squared deviation from the data, to which a penalty is added for size of the model and for species in the model that do not reach steady state.

Value

This function returns a list with elements:

Author(s)

C. Terfve. T. Cokelaer

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

See Also

gaBinaryTN, simulatorT1

Examples

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

#pre-process model

model = preprocessing(CNOlistToy, ToyModel)

#optimise

initBstring<-rep(1,length(model$reacID))
ToyT1opt<-gaBinaryT1(
	CNOlist=CNOlistToy,
	model=model,
	initBstring=initBstring,
	maxGens=100, popSize=10, verbose=FALSE)

# During the optimisation, some bits can be overwritten by your prior knowledge 
# First, you need to create a priorBitString made of NA where known bit values
# are replaced by 0 or 1
priorBitString = rep(NA, length(model$reacID))
priorBitString[1] = 0
priorBitString[2] = 1

# Second, you call the gaBinaryT1 function by providing the priorBitString
# argument:
ToyT1opt<-gaBinaryT1(CNOlist=CNOlistToy, model=model,
     initBstring=initBstring,	maxGens=10, popSize=10, verbose=FALSE, 
    priorBitString=priorBitString)

Description

This is the genetic algorithm for time point N, that should follow optimisation based on time point 1.

Replaced gaBinaryT2.

Usage

gaBinaryTN(CNOlist, model,  bStrings, sizeFac = 1e-04, NAFac = 1,
popSize = 50, pMutation = 0.5, maxTime = 60, maxGens = 500,
stallGenMax = 100, selPress = 1.2, elitism = 5, relTol = 0.1, verbose=TRUE,
priorBitString=NULL, timeIndex = NULL)

Arguments

Details

This function takes in the same input as the T1 ga, but in addition it takes in the bitstring optimised for T1, and does not take an initial bitstring. Be aware that the bitString that this function returns is one that only includes the bits that it actually looks at, i.e. the bits that were 0 in the bStringT1

Value

This function returns a list with elements:

Author(s)

C.Terfve, T. Cokelaer

See Also

getFit, simulatorT1, simulatorT2

Examples

#load data

data(CNOlistToy2,package="CellNOptR")
data(ToyModel2,package="CellNOptR")

#pre-process model

checkSignals(CNOlistToy2,ToyModel2)
model = preprocessing(CNOlistToy2, ToyModel2)

#optimise t1
ToyT1<-gaBinaryT1(
	CNOlist=CNOlistToy2,
	model=model,
	maxGens=10,
	popSize = 10,
	verbose=FALSE)

#Optimise T2

ToyT2<-gaBinaryTN(
	CNOlist=CNOlistToy2,
	model=model,
	bStrings=list(ToyT1$bString),
	maxGens=10,
	popSize = 10,
	verbose=FALSE)

Description

This function computes the value of the objective function for a model and an associated data set, as a sum of a term that computes the fit of model to data, a term that penalises the NA values produced by the model, and a term that penalises increasing size of the model.

Usage

getFit(simResults, CNOlist, model, indexList=NULL, timePoint=c("t1", "t2"),
sizeFac=1e-04, NAFac=1, nInTot, simResultsT0=NULL)

Arguments

Details

BE AWARE: contrary to what is done in the Matlab version of CellNOpt, here the simulation results are computed beforehand and the model that is input into this function is a model that has already been cut i.e. that only contains the reactions present in the optimised model (i.e.should be the same model as the one that you input into the simulator). Also, the simResults matrix is the full one as output by the simulator, i.e. it contains results for all species in the model, not only the signals

Value

This function returns a single number, the value of the objective function.

Author(s)

C. Terfve

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

See Also

gaBinaryT1, simulatorT1

Examples

#Here we will evaluate the fit of the full initial model, 
#without pre-processing or any optimisation

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=FALSE)
ToyFields4Sim<-prep4sim(ToyModel)
simResults<-simulatorT1(
	CNOlist=CNOlistToy,
	model=ToyModel,
	simList=ToyFields4Sim,
	indexList=indicesToy)
simResults = simResults[, indicesToy$signals]
Score<-getFit(
	simResults=simResults,
	CNOlist=CNOlistToy,
	model=ToyModel,
	timePoint="t1", 
    nInTot=length(which(ToyModel$interMat == -1))
    )

Description

This function converts a network form graph format to SIF format. The resulting table can also be saved in a SIF file.

Usage

graph2sif(graph, writeSif=FALSE, filename="Graph")

Arguments

Details

The sign of link is supposed to be encoded in the graph as the weigth of the edge (-1 negative regulation, +1 positive regulation).

Value

Author(s)

F. Eduati

See Also

model2sif, sif2graph, readSIF,

Examples

data(ToyModel,package="CellNOptR")
    sif_file = tempfile(fileext = ".sif")
    writeSIF(model = ToyModel,filename = sif_file)
    g = sif2graph(sif_file)
    
    graph2sif(graph = g,writeSif = FALSE)

Description

This function is the ilp method to be used to optimise a model by fitting to data containing one time point.

Usage

ilpBinaryT1(cnolist, 
                        model, 
                        cplexPath,
                        sizeFac = 0.0001, 
                        mipGap = 0, 
                        relGap = 0, 
                        timelimit = 3600, 
                        method = "quadratic",
                        numSolutions = 100, 
                        limitPop = 500, 
                        poolIntensity = 0, 
                        poolReplace = 2)

Arguments

Value

This function returns a list with elements:

Author(s)

E Gjerga, H Koch

References

Alexander Mitsos, Ioannis N. Melas, Paraskeuas Siminelakis, Aikaterini D. Chairakaki, Julio Saez-Rodriguez, and Leonidas G. Alexopoulos. Identifying Drug Effects via Pathway Alterations using an Integer Linear Programming Optimization Formulation on Phosphoproteomic Data. PLoS Comput Biol. 2009 Dec; 5(12): e1000591.

Examples

# Toy Exampple
data("ToyModel", package="CellNOptR")
data("CNOlistToy", package="CellNOptR")
pknmodel = ToyModel
cnolist = CNOlist(CNOlistToy)
model = preprocessing(data = cnolist, model = pknmodel, compression = TRUE, expansion = TRUE)
plotModel(model = model, CNOlist = cnolist)

# Training to data - ILP
## Not run: 
resILP = ilpBinaryT1(cnolist = cnolist, model = model)

## End(Not run)

Description

This function is the ilp method to be used to optimise a model by fitting to data for time point 2, that should follow optimisation based on time point 1.

Usage

ilpBinaryT2(cnolist, 
                        model, 
                        sizeFac = 0.0001, 
                        mipGap = 0, 
                        relGap = 0, 
                        timelimit = 3600, 
                        cplexPath, 
                        method = "quadratic",
                        numSolutions = 100, 
                        limitPop = 500, 
                        poolIntensity = 0, 
                        poolReplace = 2)

Arguments

Value

This function returns a list with elements:

Author(s)

E Gjerga, H Koch

References

Alexander Mitsos, Ioannis N. Melas, Paraskeuas Siminelakis, Aikaterini D. Chairakaki, Julio Saez-Rodriguez, and Leonidas G. Alexopoulos. Identifying Drug Effects via Pathway Alterations using an Integer Linear Programming Optimization Formulation on Phosphoproteomic Data. PLoS Comput Biol. 2009 Dec; 5(12): e1000591.

Examples

# Toy Exampple
data("ToyModel", package="CellNOptR")
data("CNOlistToy", package="CellNOptR")
pknmodel = ToyModel
cnolist = CNOlist(CNOlistToy)
model = preprocessing(data = cnolist, model = pknmodel, compression = TRUE, expansion = TRUE)
plotModel(model = model, CNOlist = cnolist)

# Training to data - ILP
## Not run: 
resILP = ilpBinaryT2(cnolist = cnolist, model = model)

## End(Not run)

Description

This function is the ilp method to be used to optimise a model by fitting to data for time point 2, that should follow optimisation based on time point 1.

Usage

ilpBinaryTN(cnolist, 
                        model, 
                        sizeFac = 0.0001, 
                        mipGap = 0, 
                        relGap = 0, 
                        timelimit = 3600, 
                        cplexPath, 
                        method = "quadratic",
                        numSolutions = 100, 
                        limitPop = 500, 
                        poolIntensity = 0, 
                        poolReplace = 2,
                        timeIndices = c(1, 2))

Arguments

Value

This function returns a list with elements:

Author(s)

E Gjerga, H Koch

References

Alexander Mitsos, Ioannis N. Melas, Paraskeuas Siminelakis, Aikaterini D. Chairakaki, Julio Saez-Rodriguez, and Leonidas G. Alexopoulos. Identifying Drug Effects via Pathway Alterations using an Integer Linear Programming Optimization Formulation on Phosphoproteomic Data. PLoS Comput Biol. 2009 Dec; 5(12): e1000591.

Examples

# Toy Exampple
data("ToyModel", package="CellNOptR")
data("CNOlistToy", package="CellNOptR")
pknmodel = ToyModel
cnolist = CNOlist(CNOlistToy)
model = preprocessing(data = cnolist, model = pknmodel, compression = TRUE, expansion = TRUE)
plotModel(model = model, CNOlist = cnolist)

# Training to data - ILP
## Not run: 
resILP = ilpBinaryTN(cnolist = cnolist, model = model)

## End(Not run)

Description

This function finds the indices, in the model fields, of the species that are measured/inhibited/stimulated. It looks for their position in model$namesSpecies which has the same order as the rows of interMat and notMat, and therefore these indexes can be used there as well.

Usage

indexFinder(CNOlist, model,verbose=FALSE)

Arguments

Value

a list with fields:

Note

For internal usage since version 1.3.28

Author(s)

C. Terfve

See Also

makeCNOlist, readSIF

Examples

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")
indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=TRUE)

Description

This is a list of functions that are part of CelNOptR but are not exposed to the end-user. It may be of interest for developers. They may have a manual associated to it. If so, you can get the documentation in a R console as usual by tying preceding the name of the function with question tag.

Details

Author(s)

T.Cokelaer

Description

This function takes as an input the file name where we write the ILP formulation together with CPLEX parameters and then solves the ILP problem.

Usage

invokeCPLEX(inputFileName, 
                        outputFileName, 
                        mipGap=mipGap, 
                        relGap = relGap, 
                        timelimit=timelimit, 
                        cplexPath = cplexPath, 
                        numSolutions = numSolutions, 
                        limitPop = limitPop, 
                        poolIntensity = poolIntensity, 
                        poolReplace = poolReplace)

Arguments

Author(s)

E Gjerga, H Koch

Description

This data object contains the model used in the package vignette, already loaded and formatted as a Model object. This is to be used with the data in "CNOListDREAM"

Usage

data(DreamModel)

Format

DreamModel is a list with fields "reacID" (character vector), "namesSpecies" (character vector), "interMat" (numerical matrix), "notMat"(numerical matrix).

Source

This data and model is extracted from the Matlab version of CellNOpt1.0 (http://www.ebi.ac.uk/saezrodriguez/software.html#CellNetOptimizer).

References

1. J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

2. Prill RJ, Marbach D, Saez-Rodriguez J, Sorger PK, Alexopoulos LG, Xue X, Clarke ND, Altan-Bonnet G, and Stolovitzky G. Towards a rigorous assessment of systems biology models: the DREAM3 challenges. PLoS One, 5(2):e9202, 2010.

Description

This function takes as input the output of readMIDAS and extracts the elements that are needed in a CNO project. Instead, please use the CNOlist class to read a MIDAS file that will be converted to a CNOlist.

Usage

makeCNOlist(dataset, subfield, verbose=TRUE)

Arguments

Details

Be aware that most of the functions in this package, including this one, expect the data to contain measurements at time 0, but these should all be equal to zero according to the normalisation procedure that should be used. Therefore, if you have one time point, the files valueSignals contains two matrices, one for t0 and one for t1.

If there are replicate rows in the MIDAS file (i.e., identical cues and identical time), this function averages the values of the measurements for these replicates.

Columns with the following tags are ignored: NOINHIB, NO-INHIB, NO-LIG, NOCYTO.

Value

a CNOlist with fields

Author(s)

C. Terfve, T. Cokelaer

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

See Also

readMIDAS, CNOlist-class

Examples

cpfile<-dir(system.file("ToyModel",package="CellNOptR"),full=TRUE)
file.copy(from=cpfile,to=getwd(),overwrite=TRUE)
dataToy<-readMIDAS(MIDASfile='ToyDataMMB.csv')
CNOlistToy<-makeCNOlist(dataset=dataToy,subfield=FALSE)

Description

Map an optimised model back onto the PKN model.

Usage

mapBack(model, PKN, bString)

Arguments

Value

Author(s)

F.Eduati

See Also

graph2sif, sif2graph, readSIF,

Description

This function receives as input a model object and converts it to a graph object made by igraph. igraph provides lots of utilities especially to write the file in different format such as GML.

Usage

model2igraph(model)

Arguments

Value

Author(s)

T. Cokelaer

See Also

graph2sif, sif2graph

Examples

data(ToyModel,package="CellNOptR")
    model2igraph(model=ToyModel)

Description

This function receives as input a model object and converts it to the cytoscape sif format. It can be used to convert either the whole model (before or after pre-processing) or the optimized one (if the corresponding bitString in provided). The resulting table can also be saved in a sif file.

Usage

model2sif(model,optimRes=NA,writeSif=FALSE, filename="Model")

Arguments

Details

All links in the model are converted in sif format that is sourceNode-tab-sign-tab-targetNode. If there are ANDs, they are converted using dummy nodes called and# (e.g. A+B=C will be A-tab-1-tab-and1; B-tab-1-tab-and1; and1-tab-1-tab-C).

Value

Author(s)

F.Eduati

See Also

graph2sif, sif2graph, readSIF,

Examples

data(ToyModel,package="CellNOptR")
    model2sif(model=ToyModel,writeSif = FALSE)

Description

This function takes in a CNOlist and does the normalisation of the data between 0 and 1, according to two different procedures (see details).

Usage

normaliseCNOlist(CNOlist, EC50Data=0.5, HillCoef=2, EC50Noise=0., detection=0,
    saturation=Inf, changeTh=0, norm2TorCtrl=NULL, mode="time", 
    options=list(rescale_negative = TRUE), verbose=FALSE)

Arguments

Details

The normalisation procedure works as follows:

1. every value that is out of the dynamic range of the equipment (as specified by the parameters detection and saturation are set to NA,

2. values are transformed to fold changes relative to the same condition at t0 (if mode="time") or the control condition (i.e. same inhibitors, no stimuli) at the same time (if mode="ctrl"),

3. the fold changes are transformed with a Hill function $\frac{x^{HillCoef}}{ ((EC50Data^{HillCoef}) + (x^{HillCoef}))}$

4. a penalty for "noisiness" is computed for each measurement as the value divided by the maximum value for that readout across all conditions and times (excluding values out of the dynamic range)

5. the noise penalty is transformed by a saturation function (for each measurement $\frac{x}{(EC50Noise+x}$ where $x=\frac{x}{\max{x}}$),

6. the noise penalty and Hilled fold changes are multiplied,

7. if the fold change is negative and bigger than ChangeTh, the resulting product is multiplied by -1, if the fold change is smaller than ChangeTh (either positive or negative), it is set to 0.

   The normalisation procedure applied here is explained in details in saez-Rodriguez et al. (2009).

   As the normalisation procedure works by computing a fold change relative to the same condition at time 0 or the control condition, if the aforementioned conditions have a value of zero (which is not expected with any common biochemical technique), then the fold change calculation will return a "NaN" value. If this is a problem for your particular case then we would suggest putting a dummy, very low value, instead of the zero, or setting that measurement to "NA" in the MIDAS file.

Value

a normalised CNOlist

Author(s)

C. Terfve

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

See Also

makeCNOlist

Examples

#Load a CNOlist

data(CNOlistToy,package="CellNOptR")

#Replace the values in the list by random values 
#(for demonstration purposes, when actually using this function you would simply load a non-normalised CNOlist)

CNOlistToy$valueSignals$t0<-matrix(
	data=runif(n=(dim(CNOlistToy$valueSignals$t0)[1]*dim(CNOlistToy$valueSignals$t0)[2]),min=0,max=400),
	nrow=dim(CNOlistToy$valueSignals$t0)[1],
	ncol=dim(CNOlistToy$valueSignals$t0)[2])
CNOlistToy$valueSignals[[2]]<-CNOlistToy$valueSignals[[1]]+matrix(
	data=runif(n=(dim(CNOlistToy$valueSignals$t0)[1]*dim(CNOlistToy$valueSignals$t0)[2]),min=0,max=100),
	nrow=dim(CNOlistToy$valueSignals$t0)[1],
	ncol=dim(CNOlistToy$valueSignals$t0)[2])

CNOlistToyN<-normaliseCNOlist(
	CNOlistToy, 
	EC50Data = 0.5, 
	HillCoef = 2, 
	EC50Noise = 0.1, 
	detection = 0, 
	saturation = Inf, 
	changeTh = 0, 
	mode = "time")

Description

Small in-silico case study used for demonstration. Use with CNOlist_ToyPB data.

Usage

data(PKN_ToyPB)

Format

An object of class list of length 4.

Description

A plot method for CNOlist.

Usage

signature(x="CNOlist"): Please see the help page for the plot.CNOlist method in the CellNOptR package

arguments

x

The CNOlist object to plot

Author(s)

T.Cokelaer

See Also

readMIDAS, makeCNOlist

Examples

showClass("CNOlist")

files<-dir(system.file("ToyModel",package="CellNOptR"),full=TRUE)
cnolist  = CNOlist(files[[1]])
# accessors:
getCues(cnolist)
getInhibitors(cnolist)
getSignals(cnolist)
getTimepoints(cnolist)
getStimuli(cnolist)
# In version 1.3.30 and above, use the plot method instead of former plotCNOlist function.
plot(cnolist)

Description

This function plots the data in a CNOlist as a matrix of plots with a row for each condition and a column for each signal, and an extra plot for each row that specifies which cues are present..

Usage

plotCNOlist(CNOlist)

Arguments

Details

This function can plot the normalised values or the un-normalised ones, it just needs a CNOlist.

Value

This function just produces a plot on your graphics window

Author(s)

C. Terfve

See Also

plotCNOlistPDF, plotCNOlistLarge, plotCNOlistLargePDF

Examples

data(CNOlistToy,package="CellNOptR")
plotCNOlist(CNOlistToy)

Description

This function plots the data in a CNOlist as a matrix of plots with a row for each condition and a column for each signal, Cues are simply represented by a number.

Usage

plotCNOlist2(cnolist, simulated_cnolist=NULL, ymin=0,ymax=1)

Arguments

Details

This function can plot either a single CNOlist, or 2 on top of each other.

Author(s)

T. Cokelaer

See Also

plotCNOlistPDF, plotCNOlistLarge, plotCNOlistLargePDF, plotCNOlist

Examples

# this data set is not an object so we need to convert it
data(CNOlistToy,package="CellNOptR")
cnolist = CNOlist(CNOlistToy)
plotCNOlist2(cnolist)

Description

This function plots the data in a CNOlist as a matrix of plots with a row for each condition and a column for each signal, and an extra plot for each row that specifies which cues are present.

Usage

plotCNOlistLarge(CNOlist,nsplit=4, newDevice=FALSE)

Arguments

Details

This function can plot normalised values or the un-normalised ones, it just needs a CNOlist. This function makes plots of CNOlists that are more readable when many conditions are present in the data. In addition to plotting the conditions divided into multiple plots, this function also plots the cues divided in two columns, one for inhibitors and one for stimuli.

Value

This function just produces plots on your graphics window.

Author(s)

C. Terfve

See Also

plotCNOlist, plotCNOlistPDF, plotCNOlistLargePDF

Examples

data(CNOlistDREAM,package="CellNOptR")
plotCNOlistLarge(CNOlistDREAM, nsplit=2)

Description

This function is a wrapper for plotCNOlistLarge, that plots the output directly in a pdf file.

Usage

plotCNOlistLargePDF(CNOlist, filename, nsplit, width=14, height=7)

Arguments

Details

This function makes plots of CNOlists that are more readable when many conditions are present in the data. In addition to plotting the conditions divided into multiple plots, this function also plots the cues divided in two columns, one for inhibitors and one for stimuli.

Value

This function doesn't return anything, it just produces a pdf file with your plots, in your current working directory.

Author(s)

C. Terfve

See Also

plotCNOlistLarge, plotCNOlist, plotCNOlistPDF

Examples

data(CNOlistDREAM,package="CellNOptR")
plotCNOlistLargePDF(CNOlistDREAM, filename="dreamData.pdf",nsplit=2)

Description

This function is a wrapper for plotCNOlist, that plots the output directly in a pdf file.

Usage

plotCNOlistPDF(CNOlist, filename)

Arguments

Value

This function doesn't return anything, it just produces a pdf file containing your plot, in your working directory.

Author(s)

C. Terfve

See Also

plotCNOlist, plotCNOlistLarge, plotCNOlistLargePDF

Examples

data(CNOlistToy,package="CellNOptR")
plotCNOlistPDF(CNOlist=CNOlistToy,filename="ToyModelGraph.pdf")

Description

This function takes in the results of an optimisation by gaBinaryT1 and plots the evolution of best fit and average fit against generations.

Usage

plotFit(optRes, filename = NULL)

Arguments

Value

This function doesn't return anything, it just produces a plot in your graphics window.

Author(s)

C. Terfve

See Also

gaBinaryT1

Examples

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

#process the model
model = preprocessing(CNOlistToy,ToyModel)

#optimise

ToyT1opt<-gaBinaryT1(
	CNOlist=CNOlistToy,
	model=model,
	maxGens=10,
	popSize=10,
	verbose=FALSE)

plotFit(optRes=ToyT1opt)

Description

This function can be used to plot a prior model network before any pre-processing step. However, additional information can be provided such as a CNOlist (see makeCNOlist and readMIDAS) or information related to the pre-processing steps (compression, NONC nodes, expansion gates). It can also be used to plot optimised model given the optimisation bitstring.

Usage

plotModel(model, CNOlist=NULL, bString=NULL, indexIntegr=NULL, signals=NULL, 
stimuli=NULL, inhibitors=NULL, NCNO=NULL, compressed=NULL, output="STDOUT",
filename=NULL, graphvizParams=list(), show=TRUE, remove_dot=TRUE, removeEmptyAnds=TRUE)

Arguments

Details

This function plots the model and also saves it in a dot file that can be processed later on. However, you can also save the plot in PNG or PDF or SVG format (one at a time).

The CNOlist argument contains the signals/stimuli/inhibitors so if you provide a CNOlist there is no need to use these arguments. If you decide to use them they will overwrite the contents of the CNOlist argument.

optimRes is the output of gaBinary. One of its field is called bString and contains a list of 0 and 1 (the optimisation is perfomed with a binary procedure). This list of 0 and 1 is then used to plot or not the edges of the model. However, you can provide a bitString made of floats (e.g., average of several bitStrings). In such case, edges will appear in gray light or dark according to the bistring value (between 0 and 1).

Value

a graph representation of the model

Note

This function depends on the Rgraphviz package.

Author(s)

T. Cokelaer

See Also

readMIDAS, readSIF, makeCNOlist, writeNetwork, writeDot, gaBinaryT1

Examples

data(CNOlistToy,package="CellNOptR")
    data(ToyModel,package="CellNOptR")
    res<-plotModel(ToyModel, CNOlist=CNOlistToy, compressed=c("TRAF6", "p38"), 
        graphvizParams=list(mode="classic", fontsize=30))

Description

This function is the equivalent of CNOPlotFits, it plots the data and the simulated values, along with an image plot that tells which cues were present. The plots are coloured according to the fit between data and simulated data.

Usage

plotOptimResults(simResults, expResults, times, namesCues, namesSignals,
valueCues, formalism="new")

Arguments

Details

The colouring of the background is done as follows: the mean absolute difference between observed and simulated values are computed, and colours are chosen based on this value: red (above 0.9), indianred1 (between O.8 and 0.9), lightpink2 (between 0.7 and 0.8), lightpink (between 0.6 and 0.7), mistyrose (between 0.5 and 0.6), palegoldenrod (between 0.4 and 0.5), palegreen (between 0.3 and 0.4), darkolivegreen3 (between 0.2 and 0.3), chartreuse3 (between 0.1 and 0.2), forestgreen (between 0 and 0.1). This function is used inside cutAndPlotResultsT1.

Value

This function doesn't return anything, it just produces a plot in your graphics window.

Author(s)

C. Terfve

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

See Also

cutAndPlotResultsT1

Examples

#We will plot the fit of the full initial model compared to the data, without any optimisation
#This is normally not done on a stand alone basis, but if you have a model and would like to visualise its output compared to your data, then this is what you should do

#load and prepare data

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")
indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=TRUE)
ToyFields4Sim<-prep4sim(ToyModel)

#simulate model

simRes<-simulatorT1(CNOlist=CNOlistToy,model=ToyModel, simList=ToyFields4Sim, indexList=indicesToy)

#format data and results

simResults<-list(t0=matrix(data=0,nrow=dim(simRes)[1],ncol=dim(simRes)[2]),t1=simRes)
expResults<-list(t0=CNOlistToy$valueSignals[[1]],t1=CNOlistToy$valueSignals[[2]])

#plot

plotOptimResults(
	simResults=simResults,
	expResults=expResults,
	times=CNOlistToy$timeSignals[1:2],
	namesCues=CNOlistToy$namesCues,
	namesSignals=CNOlistToy$namesSignals,
	valueCues=CNOlistToy$valueCues)

Description

This function plots the data and simulated values according to each experiment in CNOlist. The data is shown as black triangles and the simulation by a blue dashed line. The combination of cues is given by a panel where black denotes the presence and white the absence of the cue. The goodness-of-fit between model and data is color-coded on a continuous scale from white to red.

Usage

plotOptimResultsPan(simResults, yInterpol=NULL, xCoords=NULL,
CNOlist=CNOlist, formalism=c("ss1","ss2","ssN","dt","ode"), pdf=FALSE,
pdfFileName="", tPt=NULL,
plotParams = list(margin = 0.1, width=15, height=12, cmap_scale=1, cex=1.6,
ymin=NULL, Fac=1, rotation=0))

Arguments

Details

Depending on the logic formalism, this function is generally called from cutAndPlotResults*(). As shown in the example below however, it can plot the fit of any data and corresponding compatible model. The color denotes the goodness-of-fit, where white shows no difference between simulation and data and red is the maximum error from all conditions and readouts.

Value

This function does not return a value.

Author(s)

A. MacNamara

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

See Also

cutAndPlotResultsT1

Examples

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")
indicesToy <- indexFinder(CNOlistToy, ToyModel, verbose=TRUE)
ToyFields4Sim <- prep4sim(ToyModel)

# simulate model
simRes <- simulatorT1(CNOlist=CNOlistToy, model=ToyModel, simList=ToyFields4Sim, indexList=indicesToy)
simRes = simRes[, indicesToy$signals]

# format data and results

simResults <- list(t0=matrix(data=0, nrow=dim(simRes)[1], ncol=dim(simRes)[2]), t1=simRes)
# plot
plotOptimResultsPan(simResults,
  CNOlist=CNOlistToy,
  formalism="ss1",
  pdf=FALSE,
  tPt=10
)

Description

This is a wrapper for plotOptimResults

Usage

plotOptimResultsPDF(simResults, expResults, times, namesCues, namesSignals,
valueCues, filename, formalism="new")

Arguments

Details

The coloring of the background is done as follows: the mean absolute difference between observed and simulated values are computed, and colours are chosen based on this value: red (above 0.9), indianred1 (between O.8 and 0.9), lightpink2 (between 0.7 and 0.8), lightpink (between 0.6 and 0.7), mistyrose (between 0.5 and 0.6), palegoldenrod (between 0.4 and 0.5), palegreen (between 0.3 and 0.4), darkolivegreen3 (between 0.2 and 0.3), chartreuse3 (between 0.1 and 0.2), forestgreen (between 0 and 0.1). This function is used inside cutAndPlotResultsT1.

Value

This function doesn't return anything, it just produces a plot in a pdf document in your working directory.

Author(s)

C. Terfve

See Also

plotOptimResults, cutAndPlotResultsT1

Examples

#We will plot the fit of the full initial model compared to the data, without any optimisation
#This is normally not done on a stand alone basis, but if you have a model and would like to visualise 
#its output compared to your data, then this is what you should do

#load and prepare data

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")
indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=TRUE)
ToyFields4Sim<-prep4sim(ToyModel)

#simulate the model

simRes<-simulatorT1(CNOlist=CNOlistToy,model=ToyModel,simList=ToyFields4Sim,indexList=indicesToy)

#format the results and data as expected by plotOptimResults

simResults<-list(t0=matrix(data=0,nrow=dim(simRes)[1],ncol=dim(simRes)[2]),t1=simRes)
expResults<-list(t0=CNOlistToy$valueSignals[[1]],t1=CNOlistToy$valueSignals[[2]])

#plot

plotOptimResultsPDF(
	simResults=simResults,
	expResults=expResults,
	times=CNOlistToy$timeSignals[1:2],
	namesCues=CNOlistToy$namesCues,
	namesSignals=CNOlistToy$namesSignals,
	valueCues=CNOlistToy$valueCues, 
	filename="Toyfull.pdf")

Description

Adds to the model some fields that are used by the simulation engine

Usage

prep4sim(model)

Arguments

Details

This adds fields that are necessary for the simulation engine in a version that is extensible for constrained Fuzzy logic extension of the methods applied here (in development).

Value

this function returns a list with fields:

Note

For internal usage since version 1.3.28

Author(s)

C. Terfve, T. Cokelaer

See Also

simulatorT1

Examples

data(ToyModel,package="CellNOptR")
ToyFields4Sim<-prep4sim(ToyModel)

Description

This function performs any of the following preprocessing steps:

1. removes Non-Controllable and Non-Observables nodes

2. compress the model

3. and-gates expansion

Usage

preprocessing(data, model, cutNONC=TRUE, compression=TRUE,
    expansion=TRUE, ignoreList=NA, maxInputsPerGate=2, verbose=TRUE)

Arguments

Details

The function can apply any or none of the pre-processing steps. It returns the new model and the indices returned by indexFinder.

Value

the new model

Author(s)

T. Cokelaer

See Also

readSIF, readMIDAS, cutNONC, findNONC, compressModel, expandGates.

Examples

data(ToyModel,package="CellNOptR")
data(CNOlistToy,package="CellNOptR")
model = preprocessing(CNOlistToy, ToyModel, cutNONC=FALSE)

Description

This function takes in a CNOlist and does the normalisation of the data between 0 and 1, according to two different procedures (see details).

Usage

randomizeCNOlist(cnolist, sd=0.1, minValue=0, maxValue=1, mode="gaussian")

Arguments

Value

a noisy CNOlist

Author(s)

T. Cokelaer

Examples

data(CNOlistToyMMB, package="CellNOptR")
cnolist = CNOlistToyMMB
cnolist2 = randomizeCNOlist(cnolist, mode="uniform")

# a method called randomize is available in the CNOlist class so you could type:
cnolist2 = randomize(cnolist, mode="uniform")

Description

BND is a file format used by MaBoSS to store the boolean network definition. The reader works if the logic for the activation of the node is stated withing the parameter 'logic' of the node. An example file can be found in 'https://maboss.curie.fr/pub/example.bnd'.

Usage

readBND(filename)

Arguments

Value

CellNOpt network

Author(s)

Luis Tobalina

Examples

## Not run: 
model = readBND("https://maboss.curie.fr/pub/example.bnd")

## End(Not run)

Description

Read network from BNET file

Usage

readBNET(filename)

Arguments

Value

CellNOpt network

Author(s)

Luis Tobalina

Description

This function takes in a single argument, the name of a CSV MIDAS file containing the data, and returns a list that contains all the elements to build a CNOlist. The output of this function should be used as input for makeCNOlist.

Usage

readMIDAS(MIDASfile, verbose=TRUE)

Arguments

Details

This function does not return a CNOlist, but the output of this function can be used directly into makeCNOlist to create one. The MIDAS file format is described in Saez-Rodriguez et al. (2008).

If you have all of the readouts measured at the same series of time points, you can specify a unique DA: column which must have the format "DA:ALL".

Value

this function returns a list with fields:

Author(s)

C.Terfve

References

J. Saez-Rodriguez, A. Goldsipe, J. Muhlich, L. Alexopoulos, B. Millard, D. A. Lauffenburger, P. K. Sorger Flexible Informatics for Linking Experimental Data to Mathematical Models via DataRail. Bioinformatics, 24:6, 840-847 (2008).

See Also

makeCNOlist

Examples

cpfile<-dir(system.file("ToyModel",package="CellNOptR"),full=TRUE)
   file.copy(from=cpfile,to=getwd(),overwrite=TRUE)
   dataToy<-readMIDAS(MIDASfile='ToyDataMMB.csv')
   CNOlistToy<-makeCNOlist(dataset=dataToy,subfield=FALSE)

Description

This function reads a SBMLQual XML file where a model is stored. The XML is scanned and saved in SIF format in a temporary file. This file is read by readSIF. The returned object is therefore the output of readSIF.

Usage

readSBMLQual(filename)

Arguments

Value

a model list with fields:

Author(s)

T. Cokelaer

References

SBMLQual: qualitative models. See SBML.org for details.

Examples

## Not run: 
sif = readSBMLQual("test.xml")

## End(Not run)

Description

This function reads in a cytoscape SIF file and creates a model object that can be used in the CellNOptR procedure.

Usage

readSIF(sifFile)

Arguments

Details

The input argument is the name of a SIF file that contains a prior knowledge network. The SIF format (See http://wiki.cytoscape.org/Cytoscape_User_Manual/Network_Formats/) can be of the form:

nodeA typeA node2

or

node2 typeB node3 node4

with space or tabulations in between types and nodes. Spaces and tabulations have different meaning in the original cytoscape format. However, we do not differentiate them. Therefore names of the nodes cannot have white space inside them.

The accepted values for the types are -1 (inhibitor) or 1 (normal relation).

If there are ANDs they should be introduced as dummy nodes called "and#" (don't forget the number after "and" otherwise this won't be recognised). Please be aware that "and" nodes are not expected to be negated, i.e. there are not supposed to be !and1=xyz because that amounts to inverting the sign of all inputs of and1, which is more simply done at the inputs level.

The SIF format can also include unconnected node that is a row with a single name:

nodeA

Although there would be no error, these type of rows are ignored.

Value

a model list with fields:

Author(s)

C. Terfve

References

Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Research 2003 Nov; 13(11):2498-504.

Examples

cpfile<-dir(system.file("ToyModel",package="CellNOptR"),full=TRUE)
file.copy(from=cpfile,to=getwd(),overwrite=TRUE)
ToyModel<-readSIF(sifFile="ToyPKNMMB.sif")

Description

This function takes in a CNOlist and computes the residual error, which is the minimum error between the scaled continuous data and a binary boolean approximation of this data.

Usage

residualError(CNOlist)

Arguments

Details

Be aware that it is expected that $valueSignals[[1]] holds t0 (all signals=0) and $valueSignals[[2]] holds t1, $valueSignals[[3]] holds t2 and so on. The output is a list of residual errors at each time greater than 2. In addition, the total error is stored. For back compatibility, an additional field called t1andt2 is stored (NA is only t1 provided).

Value

a vector with named entries t1, t2, ...tn, t1andt2 and total. If only t1 is provided, t1andt2 is set to NA. The field t1andt2 may be removed in the future. Use the field called total instead.

Author(s)

C. Terfve, T. Cokelaer

See Also

makeCNOlist, normaliseCNOlist, getFit

Examples

data(CNOlistToy,package="CellNOptR")
resECNOlistToy<-residualError(CNOlistToy)

Description

This function receives as input a network in form sif format and converts it to graph format.

Usage

sif2graph(sif)

Arguments

Details

This function takes a network in sif format (tabel or file), i.e. sourceNode-tab-sign-tab-targetNode. If there are ANDs they should be introduced as dummy nodes called and# (don't forget the number after "and" otherwise this won't be recognised). Please be aware that "and" nodes are not expected to be negated, i.e. there are not supposed to be !and1=xyz because that amounts to inverting the sign of all inputs of and1, which is more simply done at the inputs level.

In the resulting graph, the sign of each link is encodes as the weigth of the edge (-1 negative regulation, +1 positive regulation).

Value

Author(s)

F.Eduati

See Also

graph2sif, model2sif, readSIF,

Examples

data(ToyModel,package="CellNOptR")
    sif_file = tempfile(fileext = ".sif")
    writeSIF(model = ToyModel,filename = sif_file)
    g = sif2graph(sif_file)

Description

This function cuts a model according to a bitstring optimised at T1, and simulates the model accordingly.

Usage

simulateTN(CNOlist, model, bStrings)

Arguments

Details

This function is a wrapper around the family of functions called simulatorT1 , T2 and TN.

Value

a matrix of simulated values, including all species in the model, i.e. to be used as input of gaBinaryTN.

Author(s)

T.Cokelaer, S.Schrier based on simulatorT1 (C.Terfve)

See Also

cutAndPlotResultsT1, simulatorT1

Examples

# This will compute the output of a random model obtained by randomly selecting
# which gates of the initial models are included.

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

simRes<-simulateTN(
	CNOlist=CNOlistToy, 
	model=ToyModel, 
	bStrings=list(round(runif(length(ToyModel$reacID)))))

Description

This is the simulator, inspired from BoolSimEngMKM in the Matlab CellNOpt, to be used on one time point simulations

Usage

simulatorT0(CNOlist, model, simList, indexList)

Arguments

Details

Differences from the BoolSimEngMKM simulator include: the valueInhibitors has not been previously flipped; the function outputs the values across all conditions for all species in the model, instead of only for the signal species. This is because then the output of this function can be used as initial values for the version of the simulator that works on time point 2 (not implemented in this version).

If you would like to compute the output of a model that contains some of the gates in the model but not all, we suggest that you use the function SimulateT1 and specify in the bStringT1 argument which gates you want to be included. Indeed, SimulateT1 is a wrapper around simulatorT1 that takes care of cutting the model for you before simulating it.

Value

This function outputs a single matrix of format similar to valueSignals in the CNOlist but that contains an output for each species in the model. This matrix is the simulated equivalent of valueSignals at time 1, if you consider only the columns given by indexSignals.

Author(s)

T. Cokelaer

References

1. J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

2. M. K. Morris, J. Saez-Rodriguez, D. Clarke, P. K. Sorger, D. A. Lauffenburger. Training Signaling Pathway Maps to Biochemical Data with Constrained Fuzzy Logic: Quantitative Analysis of Liver Cell Responses to Inflammatory Stimuli, PLoS Comp. Biol., 7(3): e1001099, 2011.

See Also

simulateTN, cutAndPlotResultsT1

Examples

#This computes the output of the full model, which is normally not done on a stand alone basis, but if you have a model and would like to visualise its output compared to your data, then this is what you should do

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=TRUE)
ToyFields4Sim<-prep4sim(ToyModel)

Sim<-simulatorT0(
	CNOlist=CNOlistToy,
	model=ToyModel,
	simList=ToyFields4Sim,
	indexList=indicesToy)

Description

This is the simulator, inspired from BoolSimEngMKM in the Matlab CellNOpt, to be used on one time point simulations

Usage

simulatorT1(CNOlist, model, simList, indexList, mode=1)

Arguments

Details

Differences from the BoolSimEngMKM simulator include: the valueInhibitors has not been previously flipped; the function outputs the values across all conditions for all species in the model, instead of only for the signal species. This is because then the output of this function can be used as initial values for the version of the simulator that works on time point 2 (not implemented in this version).

If you would like to compute the output of a model that contains some of the gates in the model but not all, we suggest that you use the function SimulateT1 and specify in the bStringT1 argument which gates you want to be included. Indeed, SimulateT1 is a wrapper around simulatorT1 that takes care of cutting the model for you before simulating it.

Value

This function outputs a single matrix of format similar to valueSignals in the CNOlist but that contains an output for each species in the model. This matrix is the simulated equivalent of valueSignals at time 1, if you consider only the columns given by indexSignals.

Author(s)

C. Terfve

References

1. J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

2. M. K. Morris, J. Saez-Rodriguez, D. Clarke, P. K. Sorger, D. A. Lauffenburger. Training Signaling Pathway Maps to Biochemical Data with Constrained Fuzzy Logic: Quantitative Analysis of Liver Cell Responses to Inflammatory Stimuli, PLoS Comp. Biol., 7(3): e1001099, 2011.

See Also

simulateTN, cutAndPlotResultsT1

Examples

#This computes the output of the full model, which is normally not done on a stand alone basis, but if you have a model and would like to visualise its output compared to your data, then this is what you should do

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=TRUE)
ToyFields4Sim<-prep4sim(ToyModel)

Sim<-simulatorT1(
	CNOlist=CNOlistToy,
	model=ToyModel,
	simList=ToyFields4Sim,
	indexList=indicesToy)

Description

This is a the simulator at TN using a C implementation. The computation relies on the time TN-1. T1 is a special case that is solved by using simulatorT2.

Usage

simulatorTN(simResultsPrev, CNOlist, model, simList, indexList, timeIndex=3)

Arguments

Details

Differences from the BoolSimEngMKM simulator include: the valueInhibitors has not been previously flipped; the function outputs the values across all conditions for all species in the model, instead of only for the signal species. This is because then the output of this function can be used as initial values for the version of the simulator that works on time point 2 (not implemented in this version).

If you would like to compute the output of a model that contains some of the gates in the model but not all, we suggest that you use the function SimulateT1 and specify in the bStringT1 argument which gates you want to be included. Indeed, SimulateT1 is a wrapper around simulatorT1 that takes care of cutting the model for you before simulating it.

Value

This function outputs a single matrix of format similar to valueSignals in the CNOlist but that contains an output for each species in the model. This matrix is the simulated equivalent of valueSignals at time 1, if you consider only the columns given by indexSignals.

Author(s)

T. Cokelaer, based on cSimulator (A. MacNamara)

References

1. J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

2. M. K. Morris, J. Saez-Rodriguez, D. Clarke, P. K. Sorger, D. A. Lauffenburger. Training Signaling Pathway Maps to Biochemical Data with Constrained Fuzzy Logic: Quantitative Analysis of Liver Cell Responses to Inflammatory Stimuli, PLoS Comp. Biol., 7(3): e1001099, 2011.

See Also

simulateTN, cutAndPlotResultsT1

Examples

#This computes the output of the full model, which is normally not done on a stand alone basis, but if you have a model and would like to visualise its output compared to your data, then this is what you should do

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=TRUE)
ToyFields4Sim<-prep4sim(ToyModel)

#Sim<-simulatorTN(
#	CNOlist=CNOlistToy,
#	model=ToyModel,
#	simList=ToyFields4Sim,
#	indexList=indicesToy)

Description

Export a Boolean network <network> to an sbml-qual file <fileName>. This file can then be read in using other software that supports SBMLqual standards.

The function also takes a bit string as input. It cuts the model according to the values in bitstrings and write the new model object to SBMLqual.

Usage

toSBML(network, file, bitString = c(rep(1,length(network$reacID))),version=c("standard","cellnopt"))

Arguments

Details

version = "standard" exports one transition block for each node of the network. This format is the SBMLqual standard, that can be imported then with other softwares

version = "cellnopt" the exported file follows a simplified syntax, where each edge of the network is transformed to a transition block in the SBMLqual file. Can be later imported to CellNOptR again.

Author(s)

Francesco Ceccarelli

Examples

data(ToyModel,package="CellNOptR")
  toSBML(ToyModel,file = tempfile())

Description

This data object contains the Toy model from the package vignette, already loaded and formatted as a Model object.

Usage

data(ToyModel)

Format

ToyModel is a list with fields "reacID" (character vector), "namesSpecies" (character vector), "interMat" (numerical matrix), "notMat"(numerical matrix).

Source

This data and model is extracted from the Matlab version of CellNOpt1.0 (http://www.ebi.ac.uk/saezrodriguez/software.html#CellNetOptimizer).

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

Description

This data object contains the Toy model from the package vignette, already loaded and formatted as a Model object, and modified for the 2 time points version (a negative fedback between cJun and Jnk (!cJun=Jnk) is added).

Usage

data(ToyModel)

Format

ToyModel is a list with fields "reacID" (character vector), "namesSpecies" (character vector), "interMat" (numerical matrix), "notMat"(numerical matrix).

Source

This data and model is extracted from the Matlab version of CellNOpt1.0 (http://www.ebi.ac.uk/saezrodriguez/software.html#CellNetOptimizer).

References

J. Saez-Rodriguez, L. G. Alexopoulos, J. Epperlein, R. Samaga, D. A. Lauffenburger, S. Klamt and P. K. Sorger. Discrete logic modeling as a means to link protein signaling networks with functional analysis of mammalian signal transduction, Molecular Systems Biology, 5:331, 2009.

Description

This function takes as input the integer variables assigned to each of the elements in the network and data in a CNOlist object in order to produce the set boundaries for each integer variable.

Usage

write_bounds(model,
             midasTreatmentPart,
             y_vector,
             binary_variables)

Arguments

Author(s)

E Gjerga, H Koch

References

Alexander Mitsos, Ioannis N. Melas, Paraskeuas Siminelakis, Aikaterini D. Chairakaki, Julio Saez-Rodriguez, and Leonidas G. Alexopoulos. Identifying Drug Effects via Pathway Alterations using an Integer Linear Programming Optimization Formulation on Phosphoproteomic Data. PLoS Comput Biol. 2009 Dec; 5(12): e1000591.

Description

This function takes as input the integer variables assigned to each of the elements in the network and data in a CNOlist object in order to produce the set of constraints needed to be optimized by the ILP solver.

Usage

write_constraints(model,
                              midasExperimentPart,
                              midasTreatmentPart,
                              reaction_sets,
                              y_vector, 
                              midas, 
                              binary_variables)

Arguments

Author(s)

E Gjerga, H Koch

References

Alexander Mitsos, Ioannis N. Melas, Paraskeuas Siminelakis, Aikaterini D. Chairakaki, Julio Saez-Rodriguez, and Leonidas G. Alexopoulos. Identifying Drug Effects via Pathway Alterations using an Integer Linear Programming Optimization Formulation on Phosphoproteomic Data. PLoS Comput Biol. 2009 Dec; 5(12): e1000591.

Description

This function writes a model to a Graphviz dot file with encoded features such as edge weight and nodes status (see details).

Usage

writeDot(dotNodes,dotMatrix,model,filename)

Arguments

Details

This function is not to be used on its own, it should be used internally to writeNetwork or writeScaffold. For the colouring of the nodes, nodes that are both stimulated and inhibited or any other combination, only one colour per category is used, and the following order of priority for the colours is used: signals prime over inhibited nodes which primes over stimulated nodes which primes over non-controllable/non-observable nodes, which primes over compressed. Nodes that are neither of those have a black contour, stimulated nodes are green, inhibited are red, measure are blue, compressed and non-controllable/non-observable nodes are black and dashed. Edges are coloured according to time stamp in the optimal model (green=t1, blue=t1 and/or t2, grey=neither); on the scaffold, the strokes of the edges reflects the weights in the models within reltol (i.e. for each edge, the weight is the frequency with which it appeared among the models within the relative tolerance boundaries around the best solution).

Value

This function does not have any output, it just writes a dot file in your working directory.

Author(s)

C. Terfve

References

Emden R. Gansner , Stephen C. North. An Open Graph Visualization System and Its Applications to Software Engineering. Software - Practice and Experience (1999)

See Also

writeNetwork, writeScaffold

Examples

#load data

data(CNOlistToy,package="CellNOptR")
data(ToyModel,package="CellNOptR")

#pre-process model

indicesToy<-indexFinder(CNOlistToy,ToyModel,verbose=TRUE)
ToyNCNOindices<-findNONC(ToyModel,indicesToy,verbose=TRUE)
ToyNCNOcut<-cutNONC(ToyModel,ToyNCNOindices)
indicesToyNCNOcut<-indexFinder(CNOlistToy,ToyNCNOcut)
ToyNCNOcutComp<-compressModel(ToyNCNOcut,indicesToyNCNOcut)
indicesToyNCNOcutComp<-indexFinder(CNOlistToy,ToyNCNOcutComp)
ToyNCNOcutCompExp<-expandGates(ToyNCNOcutComp)

#optimise

ToyFields4Sim<-prep4sim(ToyNCNOcutCompExp)
initBstring<-rep(1,length(ToyNCNOcutCompExp$reacID))
ToyT1opt<-gaBinaryT1(
	CNOlist=CNOlistToy,
	model=ToyNCNOcutCompExp,
	initBstring=initBstring, 
	maxGens=2,
	popSize=5, 
	verbose=TRUE)

#write network

writeNetwork(
	modelOriginal=ToyModel,
	modelComprExpanded=ToyNCNOcutCompExp,
	optimResT1=ToyT1opt,
	optimResT2=NA,
	CNOlist=CNOlistToy)