Package 'ChIPanalyser'

Title: ChIPanalyser: Predicting Transcription Factor Binding Sites
Description: ChIPanalyser is a package to predict and understand TF binding by utilizing a statistical thermodynamic model. The model incorporates 4 main factors thought to drive TF binding: Chromatin State, Binding energy, Number of bound molecules and a scaling factor modulating TF binding affinity. Taken together, ChIPanalyser produces ChIP-like profiles that closely mimic the patterns seens in real ChIP-seq data.
Authors: Patrick C.N.Martin & Nicolae Radu Zabet
Maintainer: Patrick C.N. Martin <[email protected]>
License: GPL-3
Version: 1.29.0
Built: 2024-10-30 04:50:27 UTC
Source: https://github.com/bioc/ChIPanalyser

Help Index

ChIPanalyser: Predicting Transcription Factor Binding Sites

Description

ChIPanalyser is a package to predict and understand TF binding by utilizing a statistical thermodynamic model. The model incorporates 4 main factors thought to drive TF binding: Chromatin State, Binding energy, Number of bound molecules and a scaling factor modulating TF binding affinity. Taken together, ChIPanalyser produces ChIP-like profiles that closely mimic the patterns seens in real ChIP-seq data.

Details

The DESCRIPTION file:

Package: ChIPanalyser
Type: Package
Title: ChIPanalyser: Predicting Transcription Factor Binding Sites
Authors@R: c( person(c("Patrick", "CN"), "Martin", role=c("cre", "aut"), email="[email protected]"), person(c("Nicolea","Radu"), "Zabet", role="aut"))
Version: 1.29.0
Date: 2017-09-01
Author: Patrick C.N.Martin & Nicolae Radu Zabet
Maintainer: Patrick C.N. Martin <[email protected]>
Citation: Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.
Description: ChIPanalyser is a package to predict and understand TF binding by utilizing a statistical thermodynamic model. The model incorporates 4 main factors thought to drive TF binding: Chromatin State, Binding energy, Number of bound molecules and a scaling factor modulating TF binding affinity. Taken together, ChIPanalyser produces ChIP-like profiles that closely mimic the patterns seens in real ChIP-seq data.
License: GPL-3
Collate: 2AllS4Class_ProfileParameters.R 3AllGenerics.R 4AllMethods.R AllInitialize.R AllShowMethods.R computeChIPProfile.R computeOccupancy.R computeOptimal.R computePWMScore.R computeGenomeWide.R parallelInternalFunctionsDev.R GenomicProfileGenericFunctions.R plotOccupancy.R plotOptimalHeatMapDev.R DataPreprocessingDev.R DataPreprocessingGenericFunctionsDev.R profileAccuracyEstimateDev.R GAAnalysis.R GAGeneric.R
Depends: R (>= 3.5.0),GenomicRanges, Biostrings, BSgenome, RcppRoll, parallel
Imports: methods, IRanges, S4Vectors,grDevices,graphics,stats,utils,rtracklayer,ROCR, BiocManager,GenomeInfoDb,RColorBrewer
Suggests: BSgenome.Dmelanogaster.UCSC.dm6,knitr, RUnit, BiocGenerics
Encoding: UTF-8
LazyData: true
biocViews: Software, BiologicalQuestion, WorkflowStep, Transcription, Sequencing, ChipOnChip, Coverage, Alignment, ChIPSeq, SequenceMatching, DataImport ,PeakDetection
VignetteBuilder: knitr
RoxygenNote: 7.2.1
Repository: https://bioc.r-universe.dev
RemoteUrl: https://github.com/bioc/ChIPanalyser
RemoteRef: HEAD
RemoteSha: 0c3209595339e38bf8d122dfe6a8284931d9b4f4

Index of help topics: 

BPFrequency             Accessor method for 'BPFrequency' slot in a
                        'genomicProfiles' object.
BPFrequency-methods     ~~ Methods for Function 'BPFrequency' ~~
BPFrequency<-           Setter method for 'BPFrequency' slot in a
                        'genomicProfiles' object.
BPFrequency<--methods   ~~ Methods for Function 'BPFrequency<-' ~~
ChIPScore-class         Class '"ChIPScore"'
ChIPanalyser            ChIPanalyser: Predicting Transcription Factor
                        Binding Sites
ChIPanalyserData        ChIPanalyserData
DNASequenceLength       Accessor method for 'DNASequenceLength' slot in
                        a 'genomicProfiles'
DNASequenceLength-methods
                        ~~ Methods for Function 'DNASequenceLength' ~~
GRList-class            Class '"GRList"'
PFMFormat               Accesor method for the 'PFMFormat' slot in a
                        'genomicProfiles' object
PFMFormat-methods       ~~ Methods for Function 'PFMFormat' ~~
PFMFormat<-             Setter method for the 'PFMFormat' slot in a
                        'genomicProfiles' object
PFMFormat<--methods     ~~ Methods for Function 'PFMFormat<-' ~~
PWMThreshold            Accessor method for the 'PWMThreshold' slot in
                        a 'parameterOptions' object
PWMThreshold-methods    ~~ Methods for Function 'PWMThreshold' ~~
PWMThreshold<-          Setter Method for the 'PWMThreshold' slot in a
                        'parameterOptions' object
PWMThreshold<--methods
                        ~~ Methods for Function 'PWMThreshold<-' ~~
PWMpseudocount          Accessor Method for a 'PWMpseudocount' slot in
                        a 'parameterOptions'
PWMpseudocount-methods
                        ~~ Methods for Function 'PWMpseudocount' ~~
PWMpseudocount<-        Setter Method for the 'pseudocount' slot in a
                        'parameterOptions' object
PWMpseudocount<--methods
                        ~~ Methods for Function 'PWMpseudocount<-' ~~
PositionFrequencyMatrix
                        Accessor method for the 'PFM' slot in a
                        'genomicProfiles' object
PositionFrequencyMatrix-methods
                        ~~ Methods for Function
                        'PositionFrequencyMatrix' ~~
PositionFrequencyMatrix<-
                        Setter method for the 'PFM' slot in a
                        'genomicProfiles' object
PositionFrequencyMatrix<--methods
                        ~~ Methods for Function
                        'PositionFrequencyMatrix<-' ~~
PositionWeightMatrix    Accessor Method for the 'PWM' slot in a
                        'genomicProfiles' object
PositionWeightMatrix-methods
                        ~~ Methods for Function 'PositionWeightMatrix'
                        ~~
PositionWeightMatrix<-
                        Setter Method for the 'PositionWeightMatrix'
                        slot in a 'genomicProfiles' object
PositionWeightMatrix<--methods
                        ~~ Methods for Function
                        'PositionWeightMatrix<-' ~~
averageExpPWMScore      Accessor for 'averageExpPWMScore' slot in a
                        'genomicProfiles' object.
averageExpPWMScore-methods
                        ~~ Methods for Function 'averageExpPWMScore' ~~
backgroundSignal        Accessor method for the 'backgroundSignal' slot
                        in a 'parameterOptions' object.
backgroundSignal-methods
                        ~~ Methods for Function 'backgroundSignal' ~~
backgroundSignal<-      Setter method for 'backgroundSignal' slot in a
                        'parameterOptions'
backgroundSignal<--methods
                        ~~ Methods for Function 'backgroundSignal<-' ~~
boundMolecules          Accessor methods for 'boundMolecules' slot in
                        'parameterOptions' object.
boundMolecules-methods
                        ~~ Methods for Function 'boundMolecules' ~~
boundMolecules<-        Setter method for the 'boundMolecules' slot in
                        a 'parameterOptions' object.
boundMolecules<--methods
                        ~~ Methods for Function 'boundMolecules<-' ~~
chipMean                Accessor method for 'chipMean' slot in a
                        'parameterOptions' object.
chipMean-methods        ~~ Methods for Function 'chipMean' ~~
chipMean<-              Access methods for 'chipMean' slot in
                        'parameterOptions' object.
chipMean<--methods      ~~ Methods for Function 'chipMean<-' ~~
chipSd                  Accessor method for 'chipSd' slot in a
                        'parameterOptions' object.
chipSd-methods          ~~ Methods for Function 'chipSd' ~~
chipSd<-                Setter methods for 'chipSd' slot in a
                        'parameterOptions' object.
chipSd<--methods        ~~ Methods for Function 'chipSd<-' ~~
chipSmooth              Accessor methods for 'chipSmooth' slot in a
                        'parameterOptions' object.
chipSmooth-methods      ~~ Methods for Function 'chipSmooth' ~~
chipSmooth<-            Setter method for 'chipSmooth' slot in
                        'parameterOptions' object.
chipSmooth<--methods    ~~ Methods for Function 'chipSmooth<-' ~~
computeChIPProfile      Computing ChIP-seq like profiles from Occupancy
                        data.
computeGenomeWideScores
                        Computing Genome Wide scores
computeOccupancy        Compute Occupancy values from PWM Scores based
                        on model.
computeOptimal          compute Optimal Parameters
computePWMScore         Compute PWM Scores of sites above threshold.
drop                    Accessor Method for the 'drop' slot in a
                        'genomicProfiles' object.
drop-methods            ~~ Methods for Function 'drop' ~~
evolve                  Running the ChIPanalyser implementation of a
                        Genetic algorithm.
generateStartingPopulation
                        Generate Starting population for ChIPanalyser
                        Genetic algortihm
genomicProfiles         Genomic Profile object
genomicProfiles-class   Class '"genomicProfiles"'
genomicProfilesInternal-class
                        Class '"genomicProfilesInternal"'
getHighestFitnessSolutions
                        Get Highest Fitness Solutions
getTestingData          Extract testing data from ChIPscore object
getTrainingData         Extract training data from ChIPscore object
initialize-methods      ~~ Methods for Function 'initialize' ~~
lambdaPWM               Accessor Method for the 'lambdaPWM' slot in a
                        'parameterOptions' object
lambdaPWM-methods       ~~ Methods for Function 'lambdaPWM' ~~
lambdaPWM<-             Setter Method for the 'lambdaPWM' slot in a
                        'parameterOptions' object
lambdaPWM<--methods     ~~ Methods for Function 'lambdaPWM<-' ~~
loci                    Accessor Method for the 'loci' slot in a
                        'ChIPScore' object
loci-class              Class '"loci"'
loci-methods            ~~ Methods for Function 'loci' ~~
lociWidth               Accessor Method for the 'lociWidth' slot in a
                        'parameterOptions' object
lociWidth-methods       ~~ Methods for Function 'lociWidth' ~~
lociWidth<-             Setter Method for the 'lociWidth' slot in a
                        'parameterOptions' object
lociWidth<--methods     ~~ Methods for Function 'lociWidth<-' ~~
maxPWMScore             Accessor function for 'maxPWMScore' slot in a
                        'genomicProfiles' object.
maxPWMScore-methods     ~~ Methods for Function 'maxPWMScore' ~~
maxSignal               Accessor method for the 'maxSignal' slot in a
                        'parameterOptions' object.
maxSignal-methods       ~~ Methods for Function 'maxSignal' ~~
maxSignal<-             Setter method for 'maxSignal' slot in a
                        'parameterOptions' object.
maxSignal<--methods     ~~ Methods for Function 'maxSignal<-' ~~
minPWMScore             Accessor method the 'minPWMScore' slot in a
                        'genomicProfiles' object
minPWMScore-methods     ~~ Methods for Function 'minPWMScore' ~~
naturalLog              Accessor method the 'naturalLog' slot in a
                        'parameterOptions' object.
naturalLog-methods      ~~ Methods for Function 'naturalLog' ~~
naturalLog<-            Setter method for the 'naturalLog' slot in a
                        'parameterOptions' object.
naturalLog<--methods    ~~ Methods for Function 'naturalLog<-' ~~
noOfSites               Accessor Method for the 'noOfSites' slot in a
                        'parameterOptions' object
noOfSites-methods       ~~ Methods for Function 'noOfSites' ~~
noOfSites<-             Setter Method for the 'noOfSites' slot in a
                        'parameterOptions' object.
noOfSites<--methods     ~~ Methods for Function 'noOfSites<-' ~~
noiseFilter             Accessor Method for the 'noiseFilter' slot in a
                        'parameterOptions' object
noiseFilter-methods     ~~ Methods for Function 'noiseFilter' ~~
noiseFilter<-           Setter Method for the 'noiseFilter' slot in a
                        'parameterOptions' object
noiseFilter<--methods   ~~ Methods for Function 'noiseFilter<-' ~~
nos-class               Class '"nos"'
parameterOptions        parameter Options object
parameterOptions-class
                        Class '"parameterOptions"'
ploidy                  Accessor method for the 'ploidy' slot in a
                        'parameterOptions' object
ploidy-methods          ~~ Methods for Function 'ploidy' ~~
ploidy<-                Setter Method for the 'ploidy' slot in an
                        'parameterOptions' object
ploidy<--methods        ~~ Methods for Function 'ploidy<-' ~~
plotOccupancyProfile    Plot Occupancy Profiles
plotOptimalHeatMaps     Heat Map of optimal Parameters
processingChIP          Pre-processing ChIP-seq data
profileAccuracyEstimate
                        Estimating Accuracy of predicted Profiles
profiles                ~~ Methods for Function 'profiles' ~~
removeBackground        Accessor Method for the 'removeBackground' slot
                        in a 'parameterOptions' object
removeBackground-methods
                        ~~ Methods for Function 'removeBackground' ~~
removeBackground<-      Setter Method for the 'removeBackground' slot
                        in a 'parameterOptions' object
removeBackground<--methods
                        ~~ Methods for Function 'removeBackground<-' ~~
scores                  Accessor Method for the 'scores' slot in a
                        'ChIPScore' object
scores-methods          ~~ Methods for Function 'scores' ~~
searchSites             Searching function for Sites above threshold
                        and predicted ChIP-seq Profiles
setChromatinStates      setChromatinStates
show-methods            ~~ Methods for Function 'show' ~~
singleRun               singleRun
splitData               Get Training and Testing data from ChIPscore
                        objects
stepSize                Accessor method of the 'stepSize' slot in
                        'parameterOptions' object
stepSize-methods        ~~ Methods for Function 'stepSize' ~~
stepSize<-              Setter Method for the 'stepSize' slot in a
                        'parameterOptions'
stepSize<--methods      ~~ Methods for Function 'stepSize<-' ~~
strandRule              Accessor Method for the 'strandRule' slot in a
                        'parameterOptions' object
strandRule-methods      ~~ Methods for Function 'strandRule' ~~
strandRule<-            Setter method for the 'strandRule' slot in a
                        'parameterOptions' object.
strandRule<--methods    ~~ Methods for Function 'strandRule<-' ~~
whichstrand             Accessor method for the 'whichstrand' slot in a
                        'parameterOptions' object
whichstrand-methods     ~~ Methods for Function 'whichstrand' ~~
whichstrand<-           Setter method for the 'whichstrand' slot in a
                        'parameterOptions' object
whichstrand<--methods   ~~ Methods for Function 'whichstrand<-' ~~

Author(s)

Patrick C.N. Martin <[email protected]>

And

Nicolae Radu Zabet <[email protected]>

Maintainer: Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
}
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR",BPFrequency=DNASequenceSet)

chip<-processingChIP(chip,top)
# Computing Genome Wide
GenomeWide <- computeGenomeWideScores(DNASequenceSet = DNASequenceSet,
    genomicsProfiles = GPP)

#Compute PWM Scores
PWMScores <- computePWMScore(genomicsProfiles = GenomeWide,
    DNASequenceSet = DNASequenceSet,
    loci = top, chromatinState = Access)
#Compute Occupnacy
Occupancy <- computeOccupancy(genomicsProfiles = PWMScores,
    parameterOptions = OPP)

#Compute ChIP profiles
chipProfile <- computeChIPProfile(genomicProfiles = Occupancy,
     loci = top,
    parameterOptions = OPP)
#Estimating accuracy estimate
AccuracyEstimate <- profileAccuracyEstimate(genomicProfiles = chipProfile,
     ChIPScore = chip,
     parameterOptions = OPP)

Accessor for averageExpPWMScore slot in a genomicProfiles object.

Description

Extract or Access averageExpPWMScore slot in a genomicProfiles

Usage

averageExpPWMScore(object)

Arguments

object

object is a genomicProfiles

Details

As a general rule, averageExpPWMScore is computed and updated internally by computeGenomeWideScores. Idealy, this slot should not be updated by user. The averageExpPWMScore is the sum of the exponential of every PWM score for a given DNA sequence and divided by the length of the said DNA sequence (DNASequenceLength). This can either be the full length sequence or only the accessible sequence (see computeGenomeWideScores).

Value

Returns the averageExpPWMScore of a genomicProfiles when computed.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Accessing Data
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
# Building genomicProfiles object
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR")
# Extracting AllSitesAboveThreshold slot
averageExpPWMScore(GPP)

## Note this slot is now empty as nothing has yet been computed

~~ Methods for Function averageExpPWMScore ~~

Description

~~ Methods for function averageExpPWMScore ~~

Methods

signature(object = "genomicProfilesInternal")

Accessor method for the backgroundSignal slot in a parameterOptions object.

Description

Extract or access the backgroundSignal slot in a parameterOptions object.

Usage

backgroundSignal(object)

Arguments

object

object is an parameterOptions

Details

Default Value: 0

When computing computeOccupancy, a ChIP-seq background signal is used to scale Occupancy by considering both a backgroundSignal and a maxSignal. The backgroundSignal is also used to nomalise occupancies against maxOccupancy. The backgroundSignal usually comes from experimental data and is provided by user. As a general rule, if ChIP-seq data is available and will be used in computeChIPProfile , profileAccuracyEstimate or plotOccupancyProfile, it is advised to use the backgroundSignal from this data. We strongly encourage to set values when building a parameterOptions object.

Value

Returns a backgroundSignal of a parameterOptions object.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Building occupancyProfileParameters object
OPP <- parameterOptions()
#Viewing single value in object
backgroundSignal(OPP)

~~ Methods for Function backgroundSignal ~~

Description

~~ Methods for function backgroundSignal ~~

Methods

signature(object = "parameterOptions")

Setter method for backgroundSignal slot in a parameterOptions

Description

Setter method for backgroundSignal slot in a parameterOptions

Usage

backgroundSignal(object)<-value

Arguments

object

object is an parameterOptions object.

value

value is the value to be assigned to the backgroundSignal slot in parameterOptions. backgroundSignal should be a positive value. Defualt value is 0.

Details

Defualt value: 0. When computing computeOccupancy, a ChIP-seq background signal is used to scale Occupancy by considering both a backgroundSignal and a maxSignal. The backgroundSignal is also used to nomalise occupancies to maxOccupancy. The backgroundSignal usually comes from experimental data and is provided by user. As a general rule, if ChIP-seq data is available and will be used in computeChIPProfile , profileAccuracyEstimate or plotOccupancyProfile, it is advised to use the backgroundSignal from this data. We strongly encourage to set values when building a parameterOptions object.

Value

Returns a parameterOptions object with a new value assigned to the backgroundSignal slot.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Building occupancyProfileParameters object
OPP <- parameterOptions()
# Setting new value for backgroundSignal
backgroundSignal(OPP) <- 0.2
# Viewing whole object with new updated value
OPP
#Viewing single value in object
backgroundSignal(OPP)

~~ Methods for Function backgroundSignal<- ~~

Description

~~ Methods for function backgroundSignal<- ~~

Methods

backgroundSignal(object)<-value

Accessor methods for boundMolecules slot in parameterOptions object.

Description

Extract or Access boundMolecules slot in parameterOptions object.

Usage

boundMolecules(object)

Arguments

object

object is a parameterOptions object.

Details

Defaut value: 1000

When computing occupancy (computeOccupancy), a value for the number of bound Molecules to DNA is needed. This value can be updated and set in a parameterOptions object. If the number of molecules is unknown,it is possible to infer this value with computeOptimal. We strongly encourage to set values when building a parameterOptions object.

Value

Returns boundMolecules slot in parameterOptions object.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Building parameterOptions object
OPP <- parameterOptions()
#Checking single value by slot accessor
boundMolecules(OPP)

~~ Methods for Function boundMolecules ~~

Description

~~ Methods for function boundMolecules ~~

Methods

signature(object = "parameterOptions")

Setter method for the boundMolecules slot in a parameterOptions object.

Description

Setter method for the boundMolecules slot in a parameterOptions object.

Usage

boundMolecules(object)<-value

Arguments

object

object is a parameterOptions object.

value

value is a positive integer or vector of positive integers describing the number ofmolecules bound to DNA. Defaut value is 1000.

Details

Default value: 1000 When computing occupancy (computeOccupancy), a value for the number of bound Molecules to DNA is needed. This value can be updated and set in a parameterOptions object. If the number of molecules is unknown, it is possible to infer this value with computeOptimal. We strongly encourage to set values when building a parameterOptions object.

Value

Returns a parameterOptions object with an updated value for boundMolecules.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Building parameterOptions object
OPP <- parameterOptions()
# Setting new boundMolecules value in OPP
boundMolecules(OPP) <- 5000
#Checking value in whole object
OPP
#Checking single value by slot accessor
boundMolecules(OPP)

~~ Methods for Function boundMolecules<- ~~

Description

~~ Methods for function boundMolecules<- ~~

Methods

signature(object = "parameterOptions", value = "vector")

Accessor method for BPFrequency slot in a genomicProfiles object.

Description

Extract or Access BPFrequency slot in a genomicProfiles object.

Usage

BPFrequency(object)

Arguments

object

object is a genomicProfiles

Details

Default value is c(0.25,0.25,0.25,0.25) When generating a Postion Weight Matrix from a Position Frequency Matrix, the probability of occurrence of each base pair (Base Pair Frequency) is necessary (as originally described by Gary Stormo). It is possible to set custom values for BPFrequency with a vector of length 4 containing the probability of occurrence of each base pair (A,C,G,T) in order. If Base pair frequency is unknown, BPFrequency will compute base pair frequency from a DNA sequence. The nature of this sequence can be a BSgenome or a DNAStringSet. In order to decrease run time, it is advised to use DNAStringSet

Value

Returns BPFrequency slot in genomicProfiles object.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
# Building genomicProfiles object
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR")
#Extracting BPFrequency slot
BPFrequency(GPP)

~~ Methods for Function BPFrequency ~~

Description

~~ Methods for function BPFrequency ~~

Methods

signature(object = "genomicProfilesInternal")

Setter method for BPFrequency slot in a genomicProfiles object.

Description

Setter method for BPFrequency slot in a genomicProfiles object. If base pair frequency is unknown, BPFrequency will compute base pair frequency from a DNA sequence.

Usage

BPFrequency(object)<-value

Arguments

object

object is a genomicProfiles object.

value

value can three different objects:

A vector of length 4 containing the probability of occurrence of each base pair (A,C,G,T) in order. Default value is c(0.25,0.25,0.25,0.25).

A BSgenome of the organism of interest. The base pair frequency will automatically be computed and updated in genomicProfiles.

A DNAStringSet of the organisme of interest. The base pair frequency will automatically be computed and updated in genomicProfiles (Prefered method).

Details

Default value is c(0.25,0.25,0.25,0.25) When generating a Postion Weight Matrix from a Position Frequency Matrix, the probability of occurrence of each base pair (Base Pair Frequency) is necessary (as originally described by Gary Stormo). It is possible to set custom values for BPFrequency with a vector of length 4 containing the probability of occurrence of each base pair (A,C,G,T) in order. If Base pair frequency is unknown, BPFrequency will compute base pair frequency from a DNA sequence when building a genomicProfiles object. The nature of this sequence can be aBSgenome object or a DNAStringSet. In order to decrease run time, it is advised to use DNAStringSet.

Value

Returns a genomicProfiles object with an updated value for BPFrequency.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
    }
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
# Building genomicProfiles object
GPP <- genomicProfiles(PFM=PFM,PFMFormat = "JASPAR", BPFrequency=DNASequenceSet)
# Updating BPFrequency
## !! Note!! BPFrequency is used to compute PWM from PFM
## IF updated after building GPP, then it will not influence PWM
## Advised to build with BPFrequency directly
BPFrequency(GPP) <- DNASequenceSet
BPFrequency(GPP) <- c(0.25,0.25,0.25,0.25)

~~ Methods for Function BPFrequency<- ~~

Description

~~ Methods for function BPFrequency<- ~~

Methods

signature(object = "genomicProfilesInternal", value = "DNAStringSet")
signature(object = "genomicProfilesInternal", value = "vector")

ChIPanalyserData

Description

ChIPanalyserData is derived from real biological data. The source organism is Drosophila melanogaster. The data can be described as genomic data as it contains DNA sequences, loci, genetic information, DNA accessibility data and ChIP-seq data.

Usage

data(ChIPanalyserData)

Format

  1. Accessis GRanges containing DNA Accesibility data for the sequences described above.

  2. csis GRanges containing Chromatin State data for the sequences described above.

  3. topis GRanges containing a locus of interest. In this case eve strip Locus on chromosome 2R in Drosophila melanogaster

  4. chipis a GRanges containing ChIP score of the eve strip locus in Drosophila melanogaster.

  5. geneRefis a GRanges containing UCSC gene reference information

Value

Returns a set of Rdata objects as described above.

Source

Transcription Factor PFM: Berkeley Drosophila Transcription Network Project (bdtnp.lbl.gov)

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

data(ChIPanalyserData)

Accessor method for chipMean slot in a parameterOptions object.

Description

Accessor method for chipMean slot in a parameterOptions object.

Usage

chipMean(object)

Arguments

object

object is a parameterOptions

Details

Default vlaue : 150 When computing ChIP-seq like profiles (computeChIPProfile, the occupancy values given by computeOccupancy are transformed into ChIP-seq like profiles. The average size of a ChIP-seq peak was described by Kaplan (Kaplan et al. , 2011). It is advised to use the average width of ChIP peaks from actual ChIP-seq data. We strongly encourage to set values when building a parameterOptions object.

Value

Returns chipMean slot from a parameterOptions object.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Kaplan T.,Li X.-Y.,Sabo P.J.,Thomas S.,Stamatoyannopoulos J.A., Biggin M.D., EisenM.B. Quantitative models of the mechanisms that control genome-wide patterns of transcription factor binding during early Drosophila development, PLoS Genet.,2011, vol. 7 pg. e1001290

Examples

# Building parameterOptions object
OPP <- parameterOptions()
#Accessing chipMean slot in OPP
chipMean(OPP)

~~ Methods for Function chipMean ~~

Description

~~ Methods for function chipMean ~~

Methods

chipMean(object)

Access methods for chipMean slot in parameterOptions object.

Description

Access methods for chipMean slot in parameterOptions object.

Usage

chipMean(object)<-value

Arguments

object

object is a parameterOptions object.
value

value is a positive numeric value that will be assigned to the chipMean slot. chipMean describes the average size of a ChIP-seq peak in base pairs.

Details

Default vlaue : 150 When computing ChIP-seq like profiles (computeChIPProfile, the occupancy values given by computeOccupancy are transformed into ChIP-seq like profiles. The average size of a ChIP-seq peak was described by Kaplan (Kaplan et al. , 2011). It is advised to use the average width of ChIP peaks from actual ChIP-seq data. We strongly encourage to set values when building a parameterOptions object.

Value

Returns a parameterOptions object with an updated value for chipMean slot.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Kaplan T.,Li X.-Y.,Sabo P.J.,Thomas S.,Stamatoyannopoulos J.A., Biggin M.D.,EisenM.B. Quantitative models of the mechanisms that control genome-wide patterns of transcription factor binding during early Drosophila development, PLoS Genet.,2011, vol. 7 pg. e1001290

Examples

# Building parameterOptions object
OPP <- parameterOptions()
# Setting new value for slot
chipMean(OPP) <- 250

~~ Methods for Function chipMean<- ~~

Description

~~ Methods for function chipMean<- ~~

Methods

chipMean(object)<-value

Class "ChIPScore"

Description

ChIPScore is the result of the processingChIP function. This object contains the extracted ChIP Score from ChIP data, the loci of interest and optional paramters associated to ChIPanalyser. The loci of interest will either be user provided or the top n regions as defined by the reduce argument im processingChIP. This object has the sole purpose of aiding the storage and parsing of data and parameters.

Objects from the Class

Object of this class are created internaly and will be parsed to other objects as is.

Slots

scores:

Object of class "list" List of extracted ChIP scores

loci:

Object of class "loci" GRanges containing loci of interest

ploidy:

Object of class "numeric" Ploidy level of the organism

boundMolecules:

Object of class "vector" Number of Bound molecules to the DNA

backgroundSignal:

Object of class "numeric" ChIP background signal (average ChIP score)

maxSignal:

Object of class "numeric" max ChIP signal

lociWidth:

Object of class "numeric" Width of loci if reduce is used and no loci are provided

chipMean:

Object of class "numeric" Average ChIP peak width

chipSd:

Object of class "numeric" Standard Deviation of ChIP peak width

chipSmooth:

Object of class "vector" Smoothing window width for ChIP score

stepSize:

Object of class "numeric" Defining resolution size of ChIP like profiles (10bp = signal will be only considered every 10bp)

removeBackground:

Object of class "numeric" Signal Threshold to be removed. Default removes all negative scores

noiseFilter:

Object of class "character" Type of noise filter to be used on ChIP data.

PWMThreshold:

Object of class "numeric" Threshold of PWM scores that will be selected

strandRule:

Object of class "character" Rule to compute strand score (max, mean or sum)

whichstrand:

Object of class "character" Which strand should be used to compute PWM scores.

lambdaPWM:

Object of class "vector" Lambda value - Scaling factor to the PWM

naturalLog:

Object of class "logical" PFM to PWM conversion log transform ( natural log or log2)

noOfSites:

Object of class "nos" Number of Sites in the PWM that should be used to compute PWM scores.

PWMpseudocount:

Object of class "numeric" PWM pseudocount value for PFM to PWM conversion.

paramTag:

Object of class "character" Internal Tag - Code progression

Extends

Class "parameterOptions", directly.

Methods

.loci<-

signature(object = "ChIPScore", value = "loci"): ...

.scores<-

signature(object = "ChIPScore", value = "list"): ...

initialize

signature(.Object = "ChIPScore"): ...

loci

signature(object = "ChIPScore"): ...

scores

signature(object = "ChIPScore"): ...

show

signature(object = "ChIPScore"): ...

Author(s)

Patrick C.N. Martin

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

See Also

processingChIP

Examples

showClass("ChIPScore")

Accessor method for chipSd slot in a parameterOptions object.

Description

Access or Extract chipSd slot in a parameterOptions object.

Usage

chipSd(object)

Arguments

object

object is a parameterOptions

Details

When computing ChIP-seq like profiles (computeChIPProfile, the occupancy values given by computeOccupancy are transformed into ChIP-seq like profiles. The average size of a ChIP-seq peak was described by Kaplan (Kaplan et al. , 2011). The average peak size is subject to variation. This variation is accounted for with chipSd. It is advised to use the standard deviation of ChIP peak width from actual ChIP-seq data. We strongly encourage to set values when building a parameterOptions object.

Value

Returns a parameterOptions object with an updated value for chipSd.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Kaplan T.,Li X.-Y.,Sabo P.J.,Thomas S.,Stamatoyannopoulos J.A., Biggin M.D., Eisen M.B. Quantitative models of the mechanisms that control genome-wide patterns of transcription factor binding during early Drosophila development, PLoS Genet.,2011, vol. 7 pg. e1001290

Examples

# Building parameterOptions object
OPP <- parameterOptions()
# Accessing chipSd slot
chipSd(OPP)

~~ Methods for Function chipSd ~~

Description

~~ Methods for function chipSd ~~

Methods

chipSd(object)

Setter methods for chipSd slot in a parameterOptions object.

Description

Setter methods for chipSd slot in a parameterOptions object.

Usage

chipSd(object)<-value

Arguments

object

object is parameterOptions object.

value

value is a positive nurmeric value that will be assigned to chipSd slot. Default value is 150.

Details

When computing ChIP-seq like profiles (computeChIPProfile, the occupancy values given by computeOccupancy are transformed into ChIP-seq like profiles. The average size of a ChIP-seq peak was described by Kaplan (Kaplan et al. , 2011). The average peak size is subject to variation. This variation is accounted for with chipSd. It is advised to use the standard deviation of ChIP peak width from actual ChIP-seq data. We strongly encourage to set values when building a parameterOptions object.

Value

Returns a parameterOptions object with an updated value for chipSd.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Kaplan T.,Li X.-Y.,Sabo P.J.,Thomas S.,Stamatoyannopoulos J.A., Biggin M.D., Eisen M.B. Quantitative models of the mechanisms that control genome-wide patterns of transcription factor binding during early Drosophila development, PLoS Genet.,2011, vol. 7 pg. e1001290

Examples

# Building parameterOptions object
OPP <- parameterOptions()
# Setting new value for chipSd slot
chipSd(OPP) <-  250

~~ Methods for Function chipSd<- ~~

Description

~~ Methods for function chipSd<- ~~

Methods

chipSd(object)<-value

Accessor methods for chipSmooth slot in a parameterOptions object.

Description

Access or Extract chipSmooth slot in a parameterOptions object.

Usage

chipSmooth(object)

Arguments

object

object is a parameterOptions object.

Details

When computing ChIP-seq like (computeChIPProfile) profile from occupancy data (see computeOccupancy), the profiles are smoothed using a window of a given size. The default value is set at 250 base pairs. If chipSmooth is set to 0 then the profile will not be smoothed. We strongly encourage to set values when building a parameterOptions object.

Value

Returns the chipSmooth slot in an parameterOptions object.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Building parameterOptions object
OPP <- parameterOptions()
# Accessing chipSd slot
chipSmooth(OPP)

~~ Methods for Function chipSmooth ~~

Description

~~ Methods for function chipSmooth ~~

Methods

signature(object = "parameterOptions")

Setter method for chipSmooth slot in parameterOptions object.

Description

Setter method for chipSmooth slot in parameterOptions object.

Usage

chipSmooth(object) <- value

Arguments

object

object is a parameterOptions object.

value

value is the positive numeric value to be assigned to the chipSmooth slot in parameterOptions Default value is 250 base pairs.

Details

When computing ChIP-seq like (computeChIPProfile) profile from occupancy data (see computeOccupancy), the profiles are smoothed using a window of a given size. The default value is set at 250 base pairs.If chipSmooth is set to 0 then the profile will not be smoothed. We strongly encourage to set values when building a parameterOptions object.

Value

Returns a parameterOptions object with an updated value for chipSmooth slot.

Author(s)

Patrick C.N Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Building parameterOptions object
OPP <- parameterOptions()
# Setting new value for chipSd slot
chipSmooth(OPP) <-  250

~~ Methods for Function chipSmooth<- ~~

Description

~~ Methods for function chipSmooth<- ~~

Methods

signature(object = "parameterOptions", value = "vector")

Computing ChIP-seq like profiles from Occupancy data.

Description

computeChIPProfile compute ChIP-seq like profile from occupancy data. Occupancy data is computed using computeOccupancy.

Usage

computeChIPProfile(genomicProfiles, loci, parameterOptions = NULL,
    norm = TRUE, method = c("moving_kernel","truncated_kernel","exact"),
    peakSignificantThreshold= NULL,cores=1, verbose = TRUE)

Arguments

genomicProfiles

genomicProfiles is the result of computeOccupancy. This object should be a genomicProfiles object.

loci

loci is either a GRanges or ChIPScore object. ChIPScore-class will be the result of processingChIP. This object represents the set of Loci you are interested in analysing. If you have followed the full ChIPanalyser pipe line, you would have used the processingChIP function that would return a ChIPScore-class object containing your loci of interest. GRanges are also supported if you are only using part of the pipeline.

parameterOptions

parameterOptions is a parameterOptions object. This object is used to store the numerous paramters offered by ChIPanalyser. This argument is optional as all arguments are also parse in both ChIPScore-class and genomicProfiles objects. If you wanted to make some last minute changes, parameterOptions is the way to go. We recomend that you set your desired options before hand.

norm

norm is a logical value. If TRUE, the ChIP-seq like profile will be normalised towards maximum Occupancy. If FALSE, the profile will be left as is.

method

method is a character string of one of the following: c("moving_kernel","truncated_kernel","exact"). If set to moving_kernel, the peaks will be approximated using Rcpp (Default). If set to truncated_kernel, the peaks will be approximated however this method does not require Rcpp. If set to exact, the peaks will not be approximated.

peakSignificantThreshold

peakSignificantThreshold is a threshold at which peaks will be selected. IMPORTANT: if you select "moving_kernel" as described in method then this threshold is a numeric value describing the peak tail hight cutoff value (Default = 0.001). In the case of "truncated_kernel" and "exact", the threhsold represents a distance in base pair from the peak summit at which the peak should be cut (Default = 1250). The default is set to NULL in this function. This just means that either the value is provided bu user with the appropriate method. If not, the default will be selected depending on the method selected.

cores

cores is the number of cores that will be used to compute ChIP profiles.

verbose

verbose is a logical value. If TRUE, progress messages will be displayed in console. If FALSE, no progress messages will be dispalyed in console.

Details

computeChIPProfile converts Transcription Factor occuapncy to a profile resembling the one of a ChIP-seq profile. Internally a few paramters are required to build a ChIP like profile. These parameters are either defined and stored in a ChIPScore object (Paramters are updated based on your ChIP data ), a genomicProfiles (user defined at the start of the analysis) or a parameterOptions (if you want to update values as you go along)

Value

Returns a genomicProfiles objec containing all ChIP-seq like profile for every combination of lambdaPWM and boundMolecules provided by the user.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Extracting Data
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
    }
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
# Building genomicProfiles object
GPP <- genomicProfiles(PFM=PFM, PFMFormat="JASPAR",BPFrequency=DNASequenceSet)


# Computing Genome Wide
GenomeWide <- computeGenomeWideScores(genomicProfiles = GPP,
    DNASequenceSet = DNASequenceSet)

#Compute PWM Scores
PWMScores <- computePWMScore(genomicProfiles = GenomeWide,
     DNASequenceSet = DNASequenceSet, loci = top, chromatinState = Access)
#Compute Occupnacy
Occupancy <- computeOccupancy(genomicProfiles = PWMScores)

#Compute ChIP profiles
chipProfile <- computeChIPProfile(genomicProfiles=Occupancy,loci=top)
chipProfile

Computing Genome Wide scores

Description

computeGenomeWideScores compute the max and min PWM score over the entire genome.

Usage

computeGenomeWideScores(genomicProfiles, DNASequenceSet, chromatinState = NULL, parameterOptions = NULL, cores = 1, verbose = TRUE)

Arguments

genomicProfiles

genomicProfiles is a genomicProfiles object containing the PFM, PWM of interest.

DNASequenceSet

DNASequenceSet is a BSgenome or DNAStringSet containing the sequence of the organism of interest.

chromatinState

chromatinState is a GRanges object containing the chromatin States. This can either represent regions of accessible DNA or Chromatin state affinities.

parameterOptions

parameterOptions is a parameterOptions object containing parameters that you wish to change. The genomicProfiles object will be updated using the values assigned to parameterOptions

cores

cores is the number or cores that will be used (Numeric value - Default = 1 )

verbose

verbose is a logical value that will determine if internal progress message will be printed.

Details

computeGenomeWideScores function computes PWM scores over the entire genome (or accessible Genome if chromatin State are provided ). Genome wide scores are used to determine the maximum and minimum PWM score as well as the average exponential score. These scores will in turn be used to determine which score are above the PWM theshold. The average exponential score is an integrale part of the equation used to compute Occupancy. Using defualt settings, ChIPanalyser will only compute occupancy on the top 70% of PWM scores. This threshold can be changed. See PWMThreshold

Value

Returns a genomicsProfiles object with updated values for max score, min score and averageExpPWMScore.

Author(s)

Patrick C.N Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
      if (!requireNamespace("BiocManager", quietly=TRUE))
          install.packages("BiocManager")
      BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
      }
  library(BSgenome.Dmelanogaster.UCSC.dm6)
  DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
  # Building genomicProfiles object
  GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR", BPFrequency=DNASequenceSet)


  # Computing Genome Wide
  GenomeWide <- computeGenomeWideScores(genomicProfiles = GPP,
      DNASequenceSet = DNASequenceSet)

Compute Occupancy values from PWM Scores based on model.

Description

computeOccupancy will compute the Occupancy from PWM Scores. As described in detail in the vignette, ChIPanalyser uses PWM Scores, DNA Accessibility data, the number of bound molecules and a sclaing factor of Transcription Factor specificty. This function will compute occupancy using the values assigned to each variable.

Usage

computeOccupancy(genomicProfiles,parameterOptions = NULL,
    norm = TRUE, verbose = TRUE)

Arguments

genomicProfiles

genomicProfiles is a a genomicProfiles object resulting from computePWMScore. IT is important to use this resulting object as the occuapancy will only be computed for sites above a threshold.

parameterOptions

parameterOptions is a parameterOptions object containing the adequate values assigned to each Parameter. If not Supplied (parameterOptions = NULL), a new object will be created internally using default values.

norm

norm a logical value which determines if the occupancy should be normalised or not.

verbose

verbose a logical value which determines if progress messages are printed or not.

Details

computeOccupancy will compute the Occupancy from PWM Scores. As described in detail in the vignette, ChIPanalyser uses PWM Scores, DNA Accessibility data, the number of bound molecules and a sclaing factor of Transcription Factor specificty. This function will compute occupancy using the values assigned to each variable. It should also be noted that the parameterOptions object contains a set of parameters used to compute Occupancy (not only restricted to this ). These parameters are often dependant on real ChIP-Seq data and will influence the goodness of fit between the predicted model an real ChIP-seq data. We strongly advise that the values assigned to each parameter should be customiszed in order to increase the model ageement with real world biological data.

Value

computeOccupancy will return a genomicProfiles. The main difference will reside in the profiles slot. This slot is generally a list or GRangesList. Within these list type structures are enclosed GRanges containing the positions of site above threshold, PWMScores and Occupancy for each site. The series of GRanges will depend on the number of loci that are tested and the number of element in the list will depend on the various combinations of lambdaPWM and boundMolecules.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
    }
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)

#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR",BPFrequency=DNASequenceSet)
OPP <- parameterOptions()
# Computing Genome Wide
GenomeWide <- computeGenomeWideScores(genomicProfiles = GPP,
    DNASequenceSet = DNASequenceSet)

#Compute PWM Scores
PWMScores <- computePWMScore(genomicProfiles = GenomeWide,
    DNASequenceSet = DNASequenceSet,
    loci = top,
    chromatinState = Access)
#Compute Occupnacy
Occupancy <- computeOccupancy(genomicProfiles = PWMScores,
    parameterOptions = OPP)
Occupancy

compute Optimal Parameters

Description

ChIPanalyser contains a set of functions some of which require two parameters known as lambdaPWM and as boundMolecules. These two paramters are not always known. computeOptimal will compute these values by maximising the correlation and minimising the Mean Squared Error between a predicted ChIP-seq-like profile and a real ChIP-seq profile for a given loci.

Usage

computeOptimal(genomicProfiles,DNASequenceSet, ChIPScore,chromatinState = NULL,
    parameterOptions = NULL, optimalMethod = "all",rank=FALSE,returnAll=TRUE,
    peakMethod="moving_kernel",cores=1)

Arguments

genomicProfiles

genomicProfiles is a genomicProfiles object containing at least a Postion Frequency Matrix or a Position Weight Matrix. It is strongly advised to customize this object to increase goodness of fit of the model when compared to real ChIP-seq data.

DNASequenceSet

DNASequenceSet is a DNAStringSet or a BSgenome of the full sequence of the organism of interest.

ChIPScore

ChIPScore is a named list containing ChIP-seq enrichements for each Loci of interest. This Profile should be normalised to a base pair level. In other words, there should be an enrichement score for each base pair of a given Locus.

chromatinState

chromatinState is a GRanges object containing either accesible sites or DNA affinity scores.

parameterOptions

parameterOptions is a parameterOptions object. If this object is not provided (parameterOptions = NULL), a new object will be created internally. However, it is strongly advised to tailor this object to maximise the goodness of fit of the model when compared to ChIP-seq data.

optimalMethod

optimalMethod is a character string which determines which method for optimal parameter selection should be selected. optimalMethod can be one of the following: pearson, spearman, kendall, ks, fscore, geometric,MSE, or all. Default is set to all.

rank

rank is a logical value indicating if optimal parameters should be based on rank (parameter combination occuring the most over all regions) or avaerage score (best perfomring combination of paramters on average over all regions selected). DEFAULT = FALSE

returnAll

returnAll is a logical value indicating if all internal objects should be returned. DEFAULT = TRUE. Internal objects are the following: Occupancy Scores, ChIP like profiles, goodness of fit metrics and optimal paramters. If set to FALSE, computeOptimal will only return the optimal parameters.

peakMethod

peakMethod is a character string of one of the following: c("moving_kernel","truncated_kernel","exact"). If set to moving_kernel, the peaks will be approximated using Rcpp (Default). If set to truncated_kernel, the peaks will be approximated however this method does not require Rcpp. If set to exact, the peaks will not be approximated.

cores

cores is the number cores that will be used to compute optimal set of parameters.

Details

In order to backward infer the values of lambdaPWM and boundMolecules, it is possible to use the computeOptimal to find these parameters. It should be noted that this functions requires a ChIP-seq data input. ChIPScore (ChIP-seq data). This should be the output of the processingChIP function.

Value

computeOptimal returns a list respectivly described as the optimal set of Parameters (lambda - lambdaPWM and boundMolecules), the optimal matrix (a matrix containing accuracy estimates dependant on the parameter chosen), and finally the chosen parameter. If the parameter that was chosen was "all", then each element of this list will contain the optimal set of parameters, optimal matricies for all of the aforementioned paramters (see optimalMethod).

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
    }
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
chip<-processingChIP(chip,top)
#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR",BPFrequency=DNASequenceSet)
OPP <- parameterOptions()
#Computing Optimal set of Parameters
optimalParam <- computeOptimal(genomicProfiles = GPP,
    DNASequenceSet = DNASequenceSet,
    ChIPScore = chip,
    chromatinState = Access,
    parameterOptions = OPP,
    parameter = "all",
    peakMethod="moving_kernel")

Compute PWM Scores of sites above threshold.

Description

computePWMScore will compute and extract all sites that exhibit a PWM Score higher than a threshold. This threshold (see PWMThreshold) will determine the percentage of total sites that should NOT be considered.

Usage

computePWMScore(genomicProfiles,DNASequenceSet,
    loci = NULL, chromatinState = NULL,parameterOptions=NULL,cores=1, verbose = TRUE)

Arguments

DNASequenceSet

DNASequenceSet is a DNAStringSet or a BSgenome containing the full sequence of the organism of interest.

genomicProfiles

genomicProfiles is a genomicProfiles object resulting from the computeGenomeWideScores function.

loci

loci is a GRanges object containing the Loci of interest or a ChIPScore object result of processingChIP function.

parameterOptions

parameterOptions is a parameterOptions object containing parameters that you wish to parse/change when computing PWMScores.

chromatinState

chromatinState is a GRanges object sites of accessible DNA or DNA affinity scores.

cores

cores is the number of cores used to compute PWM Scores.

verbose

verbose is a logical value indicating if progress messages should be printed or not.

Details

After determining genome wide scores, it is possible to only compute and extract high affinity sites (in the sense that they have a high PWM Score). If a PWMThreshold is not set by user, the default value is set at 0.7. This means that 70 % of sites will NOT be selected. Only the top 30 % will be computed and extracted. If one is interested in all PWM Scores at a genome wide scale ( or accessible DNA ), this is possible by setting PWMThreshold to zero.

Value

computePWMScore will return a genomicProfiles object. The profiles slot will have been updated. This slot will now contain a GRangesList with each element being a GRanges. This GRanges will contain postion of each sites (start, end and strand) and the PWMScore associated to that site.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
    }
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
chip<-processingChIP(chip,top)
#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR",BPFrequency=DNASequenceSet)


# Computing Genome Wide
GenomeWide <- computeGenomeWideScores(DNASequenceSet = DNASequenceSet,
    genomicProfiles = GPP)

#Compute PWM Scores
PWMScores <- computePWMScore(DNASequenceSet = DNASequenceSet,
    genomicProfiles = GenomeWide,
    loci  = chip, chromatinState = Access)
PWMScores

Accessor method for DNASequenceLength slot in a genomicProfiles

Description

Accessor method for DNASequenceLength slot in a genomicProfiles

Usage

DNASequenceLength(object)

Arguments

object

object is a genomicProfiles

Details

The model on which is based ChIPanalyser requires the length of the DNA sequence used to compute scores. In this circustance, this DNA Length is the total length of the DNA of the organism of interest or the the Accessible DNA at a genome wide scale.

Value

Returns DNASequenceLength slot in a genomicProfiles object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
}
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)

#Building genomicProfiles object
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR",BPFrequency=DNASequenceSet)
# Computing Genome Wide
GenomceWide <- computeGenomeWideScores(DNASequenceSet = DNASequenceSet,
    genomicProfiles = GPP)

DNASequenceLength(GenomceWide)

~~ Methods for Function DNASequenceLength ~~

Description

~~ Methods for function DNASequenceLength ~~

Methods

signature(object = "genomicProfilesInternal")

Accessor Method for the drop slot in a genomicProfiles object.

Description

Accessor Method for the drop slot in a genomicProfiles object.

Usage

drop(object)

Arguments

object

object is a genomicProfiles object.

Details

During certain computations, it is possible that the Loci of interest do no show any overlap with accesible DNA. If this were to be the case, a warning message will appear in the console but these inaccessible Loci will be stored in this slot. It is also for these reasons that it is imperative for Loci of interest to be named (in this case, a named GRanges).

Value

Returns a character string with loci containing no accesible DNA.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)
#Loading PFM files
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR")

# Loci with no acces - a warning message will be issued
#if loci do no contain accesible DNA
# Otherwise this slot will remain empty

drop(GPP)

~~ Methods for Function drop ~~

Description

~~ Methods for function drop ~~

Methods

signature(object = "genomicProfilesInternal")

Running the ChIPanalyser implementation of a Genetic algorithm.

Description

evolve pushes a starting population to evolve in a genetic algorithm.

Usage

evolve(population,DNASequenceSet,ChIPScore,
       genomicProfiles,parameters=NULL,generations=100,mutationProbability=0.3,
       offsprings=5,chromatinState=NULL,
       method="geometric", lambda=TRUE,
       checkpoint=TRUE,
       filename=NULL, cores=1)

Arguments

population

numeric value describing the number of individuals in the starting population. Alternatively - a starting population list as returned by generateStartingPopulation. NOTE: if numeric - the parameter argument is also required.

DNASequenceSet

DNAStringSet object containing DNA sequences of interest (Extracted from BSgenome)

ChIPScore

ChIPScore object as returned by the processingChIPfunction

genomicProfiles

genomicProfiles object containing minimal information (such as the PWM)

parameters

vector or list containing each parameter that should be added to the chromosome. See generateStartingPopulation

generations

numeric describing the number of generation before the Genetic algorithm should halt.

mutationProbability

numeric descrbining the rate of mutations for each surviving individual

offsprings

numeric descrbining the number of individuals surviving to the next generation

chromatinState

GRanges object containing chromatin state information. Each state should be labled in a meta data column named "name". It is advised to use numeric values for each state name.

method

character string describing the scoring metric that should be used. ChIPanalyser offers twelve different metrics: correlation coefficients (Pearson, Spearman and Kendall), Mean Squared Error (MSE), Kolmogorov–Smirnov Distance, precision, recall, accuracy, F-score, Matthew’s correlation coefficient (MCC) and Area Under Curve Receiver Operator Characteristic (AUC ROC or just AUC)

lambda

logical describing if lambda value should be pre-computed. Setting to TRUE increases the speed of the algorithm.

checkpoint

logical describing if population parameters at each generations should be saved.

filename

character string that will serve as a prefix to the saved intermediate files.

cores

numeric describing the number of cores used to run the GA.

Details

ChIPanalyser offers a way of finding optimal solution by using a genetic algorithm. Instead of running the stadard analysis, TF binding affinities to chromatin states can be extracted via this more complex method. It should be noted that this method is better suited for the analysis of chromatin states. While the algorithm still works with simple DNA Accessibility, it would potentially take more time for accuracy minor gains.

Value

Returns a named list with three elements.

  • database saves the data frame containing all scores for each individual since generation 1

  • population saves the last population with chromosome values

  • fitestsaves the fittest individual for a given generation

Author(s)

Patrick C.N. Martin <[email protected]

Examples

library(ChIPanalyser)
data(ChIPanalyserData)
# See GA vignette for usage

Generate Starting population for ChIPanalyser Genetic algortihm

Description

generateStartingPopulation generates a starting population with random traits for each individual

Usage

generateStartingPopulation(population,parameters,names=NULL)

Arguments

population

numeric value describing the number of individuals in the starting population.

parameters

vector or list containing each parameter that should be added to the chromosome.

names

character describing names that should be added to each individual.

Details

generateStartingPopulation generates a starting poppulation to be used in the genetic algortihm implemented in ChIPanalyser. There are two main ways a starting population can be generated:

  1. by name Using names of each parameter that should be parse to each "chromosome". The possible paramters are N, lambda, PWMThreshold, CS ( DNAAffinity or DNAAccessibility also works). CS values should also contain a numeric value associated to each chromatin state you wish to parse. e.g CS1 ... CS14 This will generate a value by sampling from a set of predefined value for each paramters.

  2. by value range Using a named list (names for each parameters). Each element of the list should contain three numeric values : length of range, min value, max value. (Internally - values are parse to runif)

Value

Returns a list of individuals with a random traits

Author(s)

Patrick C.N. Martin

Examples

## by name
param <- c("N","lambda","PWMThreshold","CS1","CS2","CS3")

pop <- generateStartingPopulation(20,param, names = NULL)

# by range 
paramValue <- list(c(10,1,1000),c(10,0,5),c(10,0,0.9),c(10,0,1),c(10,0,1),c(10,0,1))

pop <- generateStartingPopulation(20,paramValue,names= param)

Genomic Profile object

Description

genomicProfiles is an S4 object serving two purposes: (i) storing internal computed data and (ii) storing paramter options. This object is parsed through the different steps of the pipeline to facilitate that parsing and changing of paramters.

Usage

genomicProfiles(..., parameterOptions = NULL, genomicProfiles = NULL, ChIPScore = NULL)

Arguments

...

Any of the user available slots in genomicProfiles.

parameterOptions

If some parameters were already previously computed or stored in a parameterOptions, parsing this object will use those values instead of the default ones.

genomicProfiles

If some parameters were already previously computed or stored in a genomicProfiles, parsing this object will use those values instead of the default ones.

ChIPScore

If some parameters were already previously computed or stored in a ChIPScore, parsing this object will use those values instead of the default ones.

Details

The genomicProfiles object serves the purpose of storing, and parsing paramters and computed data between the different steps of the pipeline. When creating a genomicProfiles object it is possible to use previously computed values by simply parsing the object to the constructor function.

Value

Returns a genomicsProfiles object with updated slots for all paramters parsed.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

See Also

genomicProfiles

parameterOptions

Examples

PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
genomicProfiles()
genomicProfiles(PFM=PFM,PFMFormat="JASPAR")

Class "genomicProfiles"

Description

genomicProfiles is an S4 object serving two purposes: (i) storing internal computed data and (ii) storing paramter options. This object is parsed through the different steps of the pipeline to facilitate that parsing and changing of paramters.

Objects from the Class

Objects can be created by calls of the form genomicProfiles(ploidy, boundMolecules, backgroundSignal, maxSignal, lociWidth, chipMean, chipSd, chipSmooth, stepSize, noiseFilter, removeBackground, lambdaPWM, PWMpseudocount, naturalLog, noOfSites, PWMThreshold, strandRule, whichstrand, PFM, PWM, PFMFormat, BPFrequency, minPWMScore, maxPWMScore, profiles, DNASequenceLength, averageExpPWMScore).

Slots

PWM:

Object of class "matrix": A Position Weight Matrix (either supplied or internally computed if PFM is provided)

PFM:

Object of class "matrix": A Position Frequency Matrix (may also be a path to file containing PFM)

PFMFormat:

Object of class "character": A character string of one of the following: raw, transfac,JASPAR or sequences

BPFrequency:

Object of class "vector": Base Pair Frequency in the genome (if a DNA sequence is provided (as a DNAStringSet or BSgenome), will be automatically computed internally). Default:c(0.25,0.25,0.25,0.25)

minPWMScore:

Object of class "vector": Lowest PWM score accros the genome (computed and updated internally)

maxPWMScore:

Object of class "vector": Highest PWM score across the genome (computed and updated internally)

profiles:

Object of class "GRList": Containins GRanges with sites above threshold and associated metrics (PWMscore and Occupancy) - Computed Internally

DNASequenceLength:

Object of class "vector": Length of the Genome (or accesible genome) - computed internally

averageExpPWMScore:

Object of class "vector": Average exponential PWM score across the genome (or accesible genome) - computed internally

ZeroBackground:

Object of class "vector": Internal background value (computed internally)

drop:

Object of class "vector": Stores Loci that do contain accesible DNA if it were to be the case (computed and updated internally)

tags:

Object of class "character" ~Internal Tags~

ploidy:

Object of class "numeric": A numeric Value descibing the ploidy of the organism. Default: 2

boundMolecules:

Object of class "vector": A vector (or single value) containing the number of bound Molecules (bound Transcription Factors): Default: 1000

backgroundSignal:

Object of class "numeric": A numeric value descibing the ChIP-seq background Signal (average signal from real ChIP seq data). Default: 0

maxSignal:

Object of class "numeric": A numeric value describing the highest ChIP-seq signal (from real ChIP-seq data). Default: 1

lociWidth:

Object of class "numeric" ~~

chipMean:

Object of class "numeric": A numeric value describing the mean width of a ChIP- seq peak. Default:150

chipSd:

Object of class "numeric": A numeric value describing the standard deviation of ChIP-seq peaks. Default: 150

chipSmooth:

Object of class "vector": A numeric value describing the width of the window used to smooth Occupancy profiles into ChIP profiles. Default:250

stepSize:

Object of class "numeric": A numeric value describing the step Size (in base pairs) between each ChIP-seq score. Default:10 (Scored every 10 base pairs)

removeBackground:

Object of class "numeric": A numeric value describing the value at which score should be removed. Defualt:0 (If negative scores then remove)

noiseFilter:

Object of class "character" ~Describes the noiseFilter method that will be applied to ChIP data (Zero, mean, median, sigmoid)~

PWMThreshold:

Object of class "numeric": Threshold at which PWM Score should be selected (only sites above threshold will be selected - between 0 and 1)

strandRule:

Object of class "character": "mean", "max" or "sum" will dertermine how strand should be handle for computing PWM Scores. Default : "max"

whichstrand:

Object of class "character": "+","-" or "+-" on which strand should PWM Score be computed. Default: "+-"

lambdaPWM:

Object of class "vector" A vector (or single value) contaning values for lambdaPWM Default:1

naturalLog:

Object of class "logical": A logical value describing if natural Log will be used to compute the PWM (if FALSE then log2 will be used). Default: TRUE

noOfSites:

Object of class "nos" A Positive integer descibing number of sites (in base pair) should be used from the PFM to compute PWM. Default =0 (Full width of binding site will be used when set to 0)

PWMpseudocount:

Object of class "numeric": A numeric value describing a PWMpseudocount for PWM computation. Default:1

paramTag:

Object of class "character" ~Internal~

Extends

Class "genomicProfilesInternal", directly. Class "parameterOptions", directly.

Methods

initialize

signature(.Object = "genomicProfiles"): ...

show

signature(object = "genomicProfiles"): ...

Author(s)

Partick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

See Also

genomicProfiles parameterOptions

Examples

showClass("genomicProfiles")

Class "genomicProfilesInternal"

Description

Non exported class. Represents the stripped down version of genomicProfiles.

Objects from the Class

Created Internally.

Slots

PWM:

Object of class "matrix" ~~

PFM:

Object of class "matrix" ~~

PFMFormat:

Object of class "character" ~~

BPFrequency:

Object of class "vector" ~~

minPWMScore:

Object of class "vector" ~~

maxPWMScore:

Object of class "vector" ~~

profiles:

Object of class "GRList" ~~

DNASequenceLength:

Object of class "vector" ~~

averageExpPWMScore:

Object of class "vector" ~~

ZeroBackground:

Object of class "vector" ~~

drop:

Object of class "vector" ~~

tags:

Object of class "character" ~~

Methods

.averageExpPWMScore<-

signature(object = "genomicProfilesInternal", value = "numeric"): ...

.DNASequenceLength<-

signature(object = "genomicProfilesInternal", value = "vector"): ...

.drop<-

signature(object = "genomicProfilesInternal", value = "vector"): ...

.generatePWM

signature(object = "genomicProfilesInternal"): ...

.maxPWMScore<-

signature(object = "genomicProfilesInternal", value = "vector"): ...

.minPWMScore<-

signature(object = "genomicProfilesInternal", value = "vector"): ...

.profiles<-

signature(object = "genomicProfilesInternal", value = "GRList"): ...

.tags

signature(object = "genomicProfilesInternal"): ...

.tags<-

signature(object = "genomicProfilesInternal", value = "character"): ...

averageExpPWMScore

signature(object = "genomicProfilesInternal"): ...

BPFrequency

signature(object = "genomicProfilesInternal"): ...

BPFrequency<-

signature(object = "genomicProfilesInternal", value = "DNAStringSet"): ...

BPFrequency<-

signature(object = "genomicProfilesInternal", value = "vector"): ...

DNASequenceLength

signature(object = "genomicProfilesInternal"): ...

drop

signature(object = "genomicProfilesInternal"): ...

maxPWMScore

signature(object = "genomicProfilesInternal"): ...

minPWMScore

signature(object = "genomicProfilesInternal"): ...

PFMFormat

signature(object = "genomicProfilesInternal"): ...

PFMFormat<-

signature(object = "genomicProfilesInternal", value = "character"): ...

PositionFrequencyMatrix

signature(object = "genomicProfilesInternal"): ...

PositionFrequencyMatrix<-

signature(object = "genomicProfilesInternal", value = "character"): ...

PositionFrequencyMatrix<-

signature(object = "genomicProfilesInternal", value = "matrix"): ...

PositionWeightMatrix

signature(object = "genomicProfilesInternal"): ...

PositionWeightMatrix<-

signature(object = "genomicProfilesInternal", value = "matrix"): ...

profiles

signature(object = "genomicProfilesInternal"): ...

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

See Also

genomicProfiles

parameterOptions

Examples

showClass("genomicProfilesInternal")

Get Highest Fitness Solutions

Description

getHighestFitnessSolutions extract best solution from a ChIPanalyser GA/evolve Run.

Usage

getHighestFitnessSolutions(population,child=2,method="geometric")

Arguments

population

Population list as output by the evolve function.

child

numeric describing the number of solution to be extracted from Population list.

method

character string describing which scoring method should be used and selected from "geometric","ks","MSE","pearson","spearman","kendall", "recall","precesion","fscore","MCC","Accuracy" or "AUC".

Details

This function only serves as a way of extracting data from the poppulation list. Ultimately - it is just a wrapper for some indexing.

Value

Return the index of the top "child" solutions.

Author(s)

Patrick C.N. Martin <[email protected]>

Examples

library(ChIPanalyser)
data(ChIPanalyserData)
# See GA vignette for usage

Extract testing data from ChIPscore object

Description

getTestingData extracts selected regions from ChIPscore object to be used as testing set.

Usage

getTestingData(ChIPscore,loci = 1)

Arguments

ChIPscore

ChIPscore object as returned by processingChIP

loci

numeric describing index of loci to be used as testing data.

Value

Returns ChIPscore object with the selected testing loci.

Author(s)

Patrick C.N. Martin <[email protected]

Examples

library(ChIPanalyser)
data(ChIPanalyserData)
# See GA vignette for usage 
test <- processingChIP(chip,top)
test <- getTestingData(test, 1:2)

Extract training data from ChIPscore object

Description

getTrainingData extracts selected regions from ChIPScore object to be used as training set.

Usage

getTrainingData(ChIPscore,loci = 1)

Arguments

ChIPscore

ChIPscore object as returned by processingChIP

loci

numeric describing index of loci to be used as training data.

Value

Returns ChIPscore object with the selected training loci.

Author(s)

Patrick C.N. Martin <[email protected]

Examples

library(ChIPanalyser)
data(ChIPanalyserData)
# See GA vignette for usage 
test <- processingChIP(chip,top)
test <- getTrainingData(test, 1:2)

Class "GRList"

Description

Virutal Class to handle multiple data types for one slot ( profiles)

Objects from the Class

A virtual Class: No objects may be created from it.

Methods

GRList-class

The purpose of this virtual classe is to store data of two different formats in one slot: GRangesList and Lists

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

showClass("GRList")

~~ Methods for Function initialize ~~

Description

~~ Methods for function initialize ~~

Methods

signature(.Object = "ChIPScore")

Initialize ChIPScore

signature(.Object = "genomicProfiles")

Initialize genomicProfiles

signature(.Object = "parameterOptions")

Initialize parameterOptions

Accessor Method for the lambdaPWM slot in a parameterOptions object

Description

Accessor Method for the lambdaPWM slot in a parameterOptions object

Usage

lambdaPWM(object)

Arguments

object

object is parameterOptions object

Details

The model underlying ChIPanalyser internally infers two paramters: number of bound molecules and lambda. Lambda represents a scaling factor for the Position weight matrix (PWM). This can be described as how well does a TF discriminate between high affinity and very high affinity sites.

Value

Returns the value assigned to the lambdaPWM slot in a parameterOptions object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(lambdaPWM=1)
#Setting new Value for lambdaPWM
lambdaPWM(GPP)

~~ Methods for Function lambdaPWM ~~

Description

~~ Methods for function lambdaPWM ~~

Methods

lambdaPWM(object)

Setter Method for the lambdaPWM slot in a parameterOptions object

Description

Setter Method for the lambdaPWM slot in a parameterOptions object

Usage

lambdaPWM(object)<-value

Arguments

object

object is parameterOptions object

value

value is the numeric value to be assigned to the lambdaPWM slot. Default set at 1.

Details

The model underlying ChIPanalyser internally infers two paramters: number of bound molecules and lambda. Lambda represents a scaling factor for the Position weight matrix (PWM). This can be described as how well does a TF discriminate between high affinity and very high affinity sites.

Value

Returns the value assigned to the lambdaPWM slot in a parameterOptions object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(lambdaPWM=1)
#Setting new Value for lambdaPWM
lambdaPWM(GPP) <- 2

~~ Methods for Function lambdaPWM<- ~~

Description

Setter method for the lambdaPWM slot in the parameterOptions

Methods

lambdaPWM(object)<-value

Accessor Method for the loci slot in a ChIPScore object

Description

Setter Method for the loci slot in a ChIPScore object

Usage

loci(object)

Arguments

object

object is ChIPScore object

Details

When using the processingChIP, this functions will return a name GRanges with the loci of interest. These loci will either result from user input or extracted from the ChIP profiles (see processingChIP and lociWidth). This functions enalbles you to extract those loci from the ChIPScore object.

Value

Returns the value assigned to the loci slot in a ChIPScore object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

chip<-processingChIP(chip,top)
loci(chip)

Class "loci"

Description

Setter for Loci of interest parsed to or extracted from the ChIPScore object

Objects from the Class

A virtual Class: No objects may be created from it.

Methods

.loci<-

signature(object = "ChIPScore", value = "loci"): ...

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

See Also

ChIPScore

Examples

showClass("loci")

~~ Methods for Function loci ~~

Description

Accessor method for the loci slot in ChIPScore

Methods

loci{Object}

Loci of interest parsed to or extracted from the ChIPScore object

Accessor Method for the lociWidth slot in a parameterOptions object

Description

Setter Method for the lociWidth slot in a parameterOptions object

Usage

lociWidth(object)

Arguments

object

object is parameterOptions object

Details

When using the processingChIP function, the provided ChIP scores will be split into bins of a given size. lociWidth determines the Size of that bin. Default is set at 20 000 bp. This means that the ChIP profiles provided will be split into bins of 20 000 bp over the entire profile provided if no loci of interest is provided.

Value

Returns the value assigned to the lociWidth slot in a parameterOptions object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(lociWidth=20000)
#Accessing new Value for lociWidth
lociWidth(GPP)

~~ Methods for Function lociWidth ~~

Description

Accessor method for the loci slot in ChIPScore

Methods

lociWidth(object)

Setting width of regions when using the reduce argument and NOT providing your own loci when using the processingChIP function.

Setter Method for the lociWidth slot in a parameterOptions object

Description

Setter Method for the lociWidth slot in a parameterOptions object

Usage

lociWidth(object)<-value

Arguments

object

object is parameterOptions object

value

value is the numeric value to be assigned to the lociWidth slot. Default set at 1.

Details

When using the processingChIP function, the provided ChIP scores will be split into bins of a given size. lociWidth determines the Size of that bin. Default is set at 20 000 bp. This mean that the ChIP profiles provided will be split into bins of 20 000 bp over the entire profile provided if no loci of interest is provided.

Value

Returns the value assigned to the lociWidth slot in a parameterOptions object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(lociWidth=20000)
#Setting new Value for lociWidth
lociWidth(GPP) <- 30000

~~ Methods for Function lociWidth<- ~~

Description

Setter method for the loci slot in ChIPScore

Methods

lociWidth(Object)<-value

Accessor function for maxPWMScore slot in a genomicProfiles object.

Description

Accessor function for maxPWMScore slot in a genomicProfiles object.

Usage

maxPWMScore(object)

Arguments

object

object is a genomicProfiles object.

Details

maxPWMScore is a numerical value that can be described as the highest PWM score computed at a genome wide scale. This value is computed and updated in the genomicProfiles object after using the computeGenomeWideScores.

Value

Returns the value of assigned to the maxPWMScore slot in a genomicProfiles object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
}
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR")

# Computing Genome Wide
GenomeWide <- computeGenomeWideScores(DNASequenceSet = DNASequenceSet,
    genomicProfiles = GPP)
maxPWMScore(GenomeWide)
## If used before computeGenomeWidePWMScore, will return NULL

~~ Methods for Function maxPWMScore ~~

Description

Accessor method for maxPWMScore

Methods

maxPWMScore(object)

Accessor method for the maxSignal slot in a parameterOptions object.

Description

Accessor method for the maxSignal slot in a parameterOptions object.

Usage

maxSignal(object)

Arguments

object

object is a parameterOptions object.

Details

In the context of ChIPanalyser, maxSignal represents the maximum normalised ChIP-Seq signal of a given Transcription factor (or DNA binding protein). Although, A default value of 1 has been assigned to this slot, we strongly recommend to tailor this value accordingly. We strongly encourage to set values when building a parameterOptions object.

Value

Returns the value assigned to the maxSignal slot in a parameterOptions object.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Building parameterOptions object
OPP <- parameterOptions()
#Setting new Value for maxSignal
maxSignal(OPP)

~~ Methods for Function maxSignal ~~

Description

Accessor method for maxSignal

Methods

maxSignal(object)

Maximum ChIP signal extracted from ChIP data (see processingChIP)

Setter method for maxSignal slot in a parameterOptions object.

Description

Setter method for maxSignal slot in a parameterOptions object.

Usage

maxSignal(object) <- value

Arguments

object

object is a parameterOptions object.

value

value is a numerical value to be assigned to the maxSignal slot.

Details

In the context of ChIPanalyser, maxSignal represents the maximum normalised ChIP-Seq signal of a given Transcription factor (or DNA binding protein). Although, A default value of 1 has been assigned to this slot, we strongly recommend to tailor this value accordingly. We strongly encourage to set values when building a parameterOptions object.

Value

Returns a parameterOptions with an updated value for maxSignal.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Building parameterOptions object
OPP <- parameterOptions()
#Setting new Value for maxSignal
maxSignal(OPP) <- 1.8

~~ Methods for Function maxSignal<- ~~

Description

Setter method for maxSignal

Methods

maxSignal(Object)<-value

Maximum ChIP signal extracted from ChIP data (see processingChIP)

Accessor method the minPWMScore slot in a genomicProfiles object

Description

Accessor method the minPWMScore slot in a genomicProfiles object

Usage

minPWMScore(object)

Arguments

object

object is a genomicProfiles object.

Details

minPWMScore can be described as the lowest PWM score computed at a genome wide scale. Although it is possible to assigne a value to minPWMScore, we strongly advise to use the value computed and assigned internally. This value is computed in the computeGenomeWideScores function.

Value

Returns the value assigned to the minPWMScore slot in a genomicProfiles object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
}
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR")

# Computing Genome Wide
GenomceWide <- computeGenomeWideScores(DNASequenceSet = DNASequenceSet,
    genomicProfiles = GPP)
minPWMScore(GenomceWide)

## If used before computeGenomeWidePWMScore, will return NULL

~~ Methods for Function minPWMScore ~~

Description

Accessor for minPWMScore

Methods

minPWMScore(object)

Minimum PWM score computed during the computeGenomeWideScores step.

Accessor method the naturalLog slot in a parameterOptions object.

Description

Accessor method the naturalLog slot in a parameterOptions object.

Usage

naturalLog(object)

Arguments

object

object is parameterOptions object.

Details

During the computation of a Postion Weight Matrix, the Position Probability Matrix (derived from a Position Frequency Matrix) is log transformed. This parameter provides whcih "log transform" will be used. If TRUE, the Natural Log will bu used (ln). If FALSE, log2 will be used. We strongly encourage to set values when building a parameterOptions object.

Value

Returns the value assigned to the naturalLog slot in a parameterOptions object.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(naturalLog=TRUE)
#Setting new Value for naturalLog
naturalLog(GPP)

~~ Methods for Function naturalLog ~~

Description

Accessor method for the naturalLog slot in a parameterOptions object.

Methods

naturalLog(object)

Setter method for the naturalLog slot in a parameterOptions object.

Description

Setter method for the naturalLog slot in a parameterOptions object.

Usage

naturalLog(object)<- value

Arguments

object

object is a parameterOptions object.

value

value is a logical value that will determine if the natural log or log2 should be used for the computation of the Position Weight Matrix.

Details

During the computation of a Postion Weight Matrix, the Position Probability Matrix (derived from a Position Frequency Matrix) is log transformed. This parameter provides whcih "log transform" will be used. If TRUE, the Natural Log will bu used (ln). If FALSE, log2 will be used. We strongly encourage to set values when building a parameterOptions object.

Value

Returns parameterOptions object with an updated value for the naturalLog slot.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
OPP <- parameterOptions(naturalLog=TRUE)
#Setting new Value for naturalLog
naturalLog(OPP) <- FALSE

~~ Methods for Function naturalLog<- ~~

Description

Setter method for the naturalLog slot in a parameterOptions object.

Methods

naturalLog(object)<-value

Accessor Method for the noiseFilter slot in a parameterOptions object

Description

Accessor Method for the noiseFilter slot in a parameterOptions object

Usage

noiseFilter(object)

Arguments

object

object is parameterOptions object

Details

Noise filtering method that should be used on ChIP-seq data. Four methods are available: Zero, Mean, Median and Sigmoid. Zero removes all ChIP-seq scores bellow zero, mean under the mean score, median under median score and sigmoid assignes a weight to each score based on a logistic regression curve. Mid point is set at 95 95 quantile of ChIP-seq scores. Below midpoint will receive a score between 0 and 1 , everything above will receive a score between 1 and 2

Value

Returns the value assigned to the noiseFilter slot in a parameterOptions object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(noiseFilter="sigmoid")
#Setting new Value for noiseFilter
noiseFilter(GPP)

~~ Methods for Function noiseFilter ~~

Description

Accessor method for noiseFilter

Methods

noiseFilter(object)

Noise Filter that will be applied to ChIP scores

Setter Method for the noiseFilter slot in a parameterOptions object

Description

Setter Method for the noiseFilter slot in a parameterOptions object

Usage

noiseFilter(object) <- value

Arguments

object

object is parameterOptions object

value

value is the value to be assigned to the noiseFilter slot (zero - mean - median -sigmoid)

Details

Noise filtering method that should be used on ChIP-seq data. Four methods are available: Zero, Mean, Median and Sigmoid. Zero removes all ChIP-seq scores bellow zero, mean under the mean score, median under median score and sigmoid assignes a weight to each score based on a logistic regression curve. Mid point is set at 95 95 quantile of ChIP-seq scores. Below midpoint will receive a score between 0 and 1 , everything above will receive a score between 1 and 2

Value

Returns the value assigned to the noiseFilter slot in a parameterOptions object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(noiseFilter="sigmoid")
#Setting new Value for noiseFilter
noiseFilter(GPP) <-"zero"

~~ Methods for Function noiseFilter<- ~~

Description

Setter method for noiseFilter

Methods

noiseFilter(object)<-value

Noise Filter that will be applied to ChIP scores

Accessor Method for the noOfSites slot in a parameterOptions object

Description

Accessor Method for the noOfSites slot in a parameterOptions object

Usage

noOfSites(object)

Arguments

object

object is parameterOptions object

Details

While computing Position Weight Matricies (PWM) from Position Frequency Matricies (PFM), it is possible to restrict the number of sites that will be used to compute the PWM. The default is set at "all". In this case, all sites will be used to compute the PWM.

Value

Returns the value assigned to the noOfSites slot in a parameterOptions object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(noOfSites="all")
#Setting new Value for naturalLog
noOfSites(GPP)

~~ Methods for Function noOfSites ~~

Description

~~ Methods for function noOfSites ~~

Methods

signature(object = "parameterOptions")

Setter Method for the noOfSites slot in a parameterOptions object.

Description

Setter Method for the noOfSites slot in a parameterOptions object.

Usage

noOfSites(object) <- value

Arguments

object

object is a parameterOptions object.

value

value is a positive integer that will be assigned to the noOfSites slot.

Details

While computing Position Weight Matricies (PWM) from Position Frequency Matricies (PFM), it is possible to restrict the number of sites that will be used to compute the PWM. The default is set at "all". In this case, all sites will be used to compute the PWM.

Value

Returns a parameterOptions object with an updated value for the noOfSites slot.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(noOfSites=0)
#Setting new Value for naturalLog
noOfSites(GPP) <- 8

~~ Methods for Function noOfSites<- ~~

Description

Setter method for noOfSites

Methods

noOfSites(object)<-"all"
noOfSites(object)<-value

Class "nos"

Description

Virtual class to handle Number of Sites

Objects from the Class

A virtual Class: No objects may be created from it.

Methods

No methods defined with class "nos" in the signature.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

showClass("nos")

parameter Options object

Description

parameterOptions is an object used to store and parse the various parameters needed throughout this analysis pipeline.

Usage

parameterOptions(ploidy = 2, boundMolecules = 1000, backgroundSignal = 0, maxSignal = 1, lociWidth = 20000, chipMean = 200, chipSd = 200, chipSmooth = 250, stepSize = 10, removeBackground = 0, noiseFilter = "zero", naturalLog = TRUE, noOfSites = "all", PWMThreshold = 0.7, strandRule = "max", whichstrand = "+-", PWMpseudocount = 1, lambdaPWM = 1)

Arguments

ploidy

ploidy: A numeric Value descibing the ploidy of the organism. Default: 2
boundMolecules

boundMolecules: A vector (or single value) containing the number of bound Molecules (bound Transcription Factors): Default: 1000
backgroundSignal

backgroundSignal: A numeric value descibing the ChIP-seq background Signal (average signal from real ChIP seq data). Default: 0
maxSignal

maxSignal: A numeric value describing the highest ChIP-seq signal (from real ChIP-seq data). Default: 1
lociWidth

lociWidth: A numeric value describing the width of the bins used to split ChIP profiles parsed to processingChIP. Default = 20000
chipMean

chipMean: A numeric value describing the mean width of a ChIP- seq peak: Default:200
chipSd

chipSd: A numeric value describing the standard deviation of ChIP-seq peaks. Default: 200

chipSmooth

chipSmooth: A numeric value describing the width of the window used to smooth Occupancy profiles into ChIP profiles. Default:250
stepSize

stepSize: A numeric value describing the step Size (in base pairs) between each ChIP-seq score. Default:10 (Scored every 10 base pairs)
removeBackground

removeBackground: A numeric value describing the value at which score should be removed. Defualt:0 (If negative scores then remove)

noiseFilter

noiseFilter: A character string of one of the following: Zero, Mean, Median, or Sigmoid. Noise filter that will be applied to the ChIP Score during the processingChIP step.

naturalLog

naturalLog: A logical value describing if natural Log will be used to compute the PWM (if FALSE then log2 will be used). Default: TRUE

noOfSites

noOfSites: A Positive integer descibing number of sites (in base pair) should be used from the PFM to compute PWM. Default =0 (Full width of binding site will be used when set to 0)

PWMThreshold

PWMThreshold: Threshold at which PWM Score should be selected (only sites above threshold will be selected - between 0 and 1)

strandRule

strandRule: ‘mean’, ‘max’ or ‘sum’ will dertermine how strand should be handle for computing PWM Scores. Default : ‘max’

whichstrand

whichstrand: ‘+’,‘-’ or ‘+-’ on which strand should PWM Score be computed. Default: ‘+-’

PWMpseudocount

PWMpseudocount: A numeric value describing a PWMpseudocount for PWM computation. Default:1

lambdaPWM

A vector (or single value) contaning values for the ScalingFactorPWM (Also known as lambda).Default:1

Details

ChIPanalyser requires a lot of parameters. parameterOptions was created with the intent of storing and parsing these numerous arguments to the different functions. All parameters in this object are optional although strongly recommend. Some parameters are extracted and updated from function along the pipeline e.g. maxSignal and backgroundSignal are extracted during the processingChIP step. These paramters will be automatically parsed. If you do not which to use them ( or any other parameter) simply parse a new parameterOptions object with your desired paramters.

Value

Returns a parameterOptions with updated values.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

See Also

genomicProfiles

Examples

#
parameterOptions(ploidy = 2, boundMolecules = 1000, backgroundSignal = 0,
    maxSignal = 1, lociWidth = 20000, chipMean = 200, chipSd = 200,
    chipSmooth = 250, stepSize = 10, removeBackground = 0, noiseFilter = "zero",
    naturalLog = TRUE, noOfSites = "all", PWMThreshold = 0.7,
    strandRule = "max", whichstrand = "+-", PWMpseudocount = 1,
    lambdaPWM = 1)

Class "parameterOptions"

Description

parameterOptions is an object used to store and parse the various parameters needed throughout this analysis pipeline.

Objects from the Class

Objects can be created by calls of the form parameterOptions(ploidy, boundMolecules, backgroundSignal, maxSignal, lociWidth, chipMean, chipSd, chipSmooth, stepSize, noiseFilter, removeBackground, lambdaPWM, PWMpseudocount, naturalLog, noOfSites, PWMThreshold, strandRule, whichstrand).

Slots

ploidy:

Object of class "numeric": A numeric Value descibing the ploidy of the organism. Default: 2

boundMolecules:

Object of class "vector": A vector (or single value) containing the number of bound Molecules (bound Transcription Factors): Default: 1000

backgroundSignal:

Object of class "numeric": A numeric value descibing the ChIP-seq background Signal (average signal from real ChIP seq data). Default: 0

maxSignal:

Object of class "numeric": A numeric value describing the highest ChIP-seq signal (from real ChIP-seq data). Default: 1

lociWidth:

Object of class "numeric": A numeric value describing bin size when splitting ChIP seq scores). Default: 20 000

chipMean:

Object of class "numeric": A numeric value describing the mean width of a ChIP- seq peak. Default:150

chipSd:

Object of class "numeric": A numeric value describing the standard deviation of ChIP-seq peaks. Default: 150

chipSmooth:

Object of class "vector": A numeric value describing the width of the window used to smooth Occupancy profiles into ChIP profiles. Default:250

stepSize:

Object of class "numeric": A numeric value describing the step Size (in base pairs) between each ChIP-seq score. Default:10 (Scored every 10 base pairs)

removeBackground:

Object of class "numeric": A numeric value describing the value at which score should be removed. Defualt:0 (If negative scores then remove)

noiseFilter:

Object of class "character" Describes noiseFilter method applied to ChIP scores

PWMThreshold:

Object of class "numeric": Threshold at which PWM Score should be selected (only sites above threshold will be selected - between 0 and 1)

strandRule:

Object of class "character": "mean", "max" or "sum" will dertermine how strand should be handle for computing PWM Scores. Default : "max"

whichstrand:

Object of class "character": "+","-" or "+-" on which strand should PWM Score be computed. Default: "+-"

lambdaPWM:

Object of class "vector" A vector (or single value) contaning values for lambdaPWM Default:1

naturalLog:

Object of class "logical": A logical value describing if natural Log will be used to compute the PWM (if FALSE then log2 will be used). Default: TRUE

noOfSites:

Object of class "nos" A Positive integer descibing number of sites (in base pair) should be used from the PFM to compute PWM. Default =0 (Full width of binding site will be used when set to 0)

PWMpseudocount:

Object of class "numeric": A numeric value describing a PWMpseudocount for PWM computation. Default:1

paramTag:

Object of class "character" ~Internal~

Methods

.paramTag

signature(object = "parameterOptions"): ...

.paramTag<-

signature(object = "parameterOptions", value = "character"): ...

.ZeroBackground

signature(object = "parameterOptions"): ...

.ZeroBackground<-

signature(object = "parameterOptions", value = "vector"): ...

backgroundSignal

signature(object = "parameterOptions"): ...

backgroundSignal<-

signature(object = "parameterOptions", value = "numeric"): ...

boundMolecules

signature(object = "parameterOptions"): ...

boundMolecules<-

signature(object = "parameterOptions", value = "vector"): ...

chipMean

signature(object = "parameterOptions"): ...

chipMean<-

signature(object = "parameterOptions", value = "numeric"): ...

chipSd

signature(object = "parameterOptions"): ...

chipSd<-

signature(object = "parameterOptions", value = "numeric"): ...

chipSmooth

signature(object = "parameterOptions"): ...

chipSmooth<-

signature(object = "parameterOptions", value = "vector"): ...

initialize

signature(.Object = "parameterOptions"): ...

lambdaPWM

signature(object = "parameterOptions"): ...

lambdaPWM<-

signature(object = "parameterOptions", value = "vector"): ...

lociWidth

signature(object = "parameterOptions"): ...

lociWidth<-

signature(object = "parameterOptions", value = "numeric"): ...

maxSignal

signature(object = "parameterOptions"): ...

maxSignal<-

signature(object = "parameterOptions", value = "numeric"): ...

naturalLog

signature(object = "parameterOptions"): ...

naturalLog<-

signature(object = "parameterOptions", value = "logical"): ...

noiseFilter

signature(object = "parameterOptions"): ...

noiseFilter<-

signature(object = "parameterOptions", value = "character"): ...

noOfSites

signature(object = "parameterOptions"): ...

noOfSites<-

signature(object = "parameterOptions", value = "character"): ...

noOfSites<-

signature(object = "parameterOptions", value = "numeric"): ...

ploidy

signature(object = "parameterOptions"): ...

ploidy<-

signature(object = "parameterOptions", value = "numeric"): ...

PWMpseudocount

signature(object = "parameterOptions"): ...

PWMpseudocount<-

signature(object = "parameterOptions", value = "numeric"): ...

PWMThreshold

signature(object = "parameterOptions"): ...

PWMThreshold<-

signature(object = "parameterOptions", value = "numeric"): ...

removeBackground

signature(object = "parameterOptions"): ...

removeBackground<-

signature(object = "parameterOptions", value = "vector"): ...

show

signature(object = "parameterOptions"): ...

stepSize

signature(object = "parameterOptions"): ...

stepSize<-

signature(object = "parameterOptions", value = "numeric"): ...

strandRule

signature(object = "parameterOptions"): ...

strandRule<-

signature(object = "parameterOptions", value = "character"): ...

whichstrand

signature(object = "parameterOptions"): ...

whichstrand<-

signature(object = "parameterOptions", value = "character"): ...

Author(s)

Partick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

See Also

genomicProfiles

Examples

showClass("parameterOptions")

Accesor method for the PFMFormat slot in a genomicProfiles object

Description

Accesor method for the PFMFormat slot in a genomicProfiles object

Usage

PFMFormat(object)

Arguments

object

object is a genomicProfiles object

Details

If loading a PositionFrequencyMatrix from a file, the format of the file should be specified. Default is raw. Please keep in mind that this argument is used when parsing the PositionFrequencyMatrix file. IF this argument is changed after building the genomicProfiles with a PositionFrequencyMatrix file, this will not influence the parsing of the file. PFMFormat can be one of the following: "raw","transfac","JASPAR" or "sequences"

Value

Returns the value assigned to the PFMFormat slot a genomicProfiles

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)
#Loading PFM files
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#Building data objects
#### THIS IS THE PREFFERED METHOD FOR SETTING PFMFormat
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR")
#Setting New value for PFMFormat
PFMFormat(GPP)

~~ Methods for Function PFMFormat ~~

Description

Accesor method for the PFMFormat slot in a genomicProfiles object

Methods

PFMFormat(object)

Setter method for the PFMFormat slot in a genomicProfiles object

Description

Setter method for the PFMFormat slot in a genomicProfiles object

Usage

PFMFormat(object) <- value

Arguments

object

object is a genomicProfiles object

value

value is character string of one of the following: "raw","transfac","JASPAR" or "sequences". If loading a PositionFrequencyMatrix from a file, the format of the file should specified. Default is JASPAR.

Details

If loading a PositionFrequencyMatrix from a file, the format of the file should be specified. Default is JASPAR. Please keep in mind that this argument is used when parsing the PositionFrequencyMatrix file. IF this argument is changed after building the genomicProfiles with a PositionFrequencyMatrix file, this will not influence the parsing of the file.

Value

Returns a genomicProfiles object with an updated value for the PFMFormat slot.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)
#Loading PFM files
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#Building data objects
#### THIS IS THE PREFFERED METHOD FOR SETTING PFMFormat
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR")
#Setting New value for PFMFormat
PFMFormat(GPP) <- "JASPAR"

~~ Methods for Function PFMFormat<- ~~

Description

Setter method for the PFMFormat slot in a genomicProfiles object

Methods

PFMFormat(object)<-value

Accessor method for the ploidy slot in a parameterOptions object

Description

Accessor method for the ploidy slot in a parameterOptions object

Usage

ploidy(object)

Arguments

object

object is a parameterOptions object

Details

Default value for ploidy is set a 2. It should be mentioned that ChIPanalyser is based on a model that also considers the ploidy of the organism of interest however this only considers simple polyploidy (or haploidy). The model does not consider hybrids such as wheat.

Value

Returns the value assigned to the ploidy slot in a parameterOptions object

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Building parameterOptions object
OPP <- parameterOptions()
#Setting new Value for maxSignal
ploidy(OPP)

~~ Methods for Function ploidy ~~

Description

Accessor method for the ploidy slot in a parameterOptions object

Methods

ploidy(object)

Setter Method for the ploidy slot in an parameterOptions object

Description

Setter Method for the ploidy slot in an parameterOptions object

Usage

ploidy(object)<- value

Arguments

object

object is a parameterOptions object

value

value is a positive integer that describes the ploidy of the organism of interest.

Details

Default value for ploidy is set a 2. It should be mentioned that ChIPanalyser is based on a model that also considers the ploidy of the organism however this only considers simple polyploidy (or haploidy). T he model does not consider hybrids such as wheat.

Value

Returns a parameterOptions object with an updated value for the ploidy slot.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Building parameterOptions object
OPP <- parameterOptions()
#Setting new Value for maxSignal
ploidy(OPP) <- 2

~~ Methods for Function ploidy<- ~~

Description

Setter Method for the ploidy slot in an parameterOptions object

Methods

ploidy(object)<-value

Plot Occupancy Profiles

Description

plotOccupancyProfile plots the predicted profiles. If provided, this functions will also plot ChIP-seq profiles, PWMScores (or Occupancy), chromatin States, Goodness of Fit estimates and gene information.

Usage

plotOccupancyProfile(predictedProfile, ChIPScore = NULL,chromatinState = NULL,
    occupancy = NULL,goodnessOfFit = NULL,PWM=FALSE,
    geneRef = NULL,addLegend = TRUE,...)

Arguments

predictedProfile

predictedProfile is a either GRanges containing the predicted profiles for one loci, all loci selected for one paramter, or all loci selected for all parameter combinations selected. (see searchSites)

ChIPScore

ChIPScore is a ChIPscore object containing ChIPscore (or a list of numeric values representing ChIP scores (Experimental ChIP))

chromatinState

chromatinState is a GRanges containing accesible DNA sites or chromatin States.

occupancy

occupancy is a GRanges or a genomicProfiles object contaning PWM scores and Occupancy ( see computeOccupancy)

goodnessOfFit

goodnessOfFit results of the profileAccuracyEstimate function.
PWM

PWM is a logical value that in the case occupancy is provided which of occupancy scores of PWM scores hsould be plotted. Default set at FALSE

geneRef

geneRef is a GRanges containing gene information on exons,introns, UTR's, enhancers or any other genetic element to be plotted.

addLegend

addLegend is a logical value defining if the legend should be added. The legend will add all elements provided. See details.

...

Any other graphical Parameter of the following : cex, cex.lab, cex.main, densityCS , densityGR , ylab, xlab, main, colPred, colChIP, colOccup, colCS, colGR, n_axis_ticks. See details.

Details

Once the predicted ChIP-seq like profiles have been computed, it is possible to plot these profiles.

This functions allows to control graphical parameters. In short:

* col = color values - exact number of colors or colors that will be used in a colorRampPalettte.

* cex = font sizes - for text, axis labels and main

* Density = fill density for chromatin state and/or geneRef blocks

Pred = predictedProfile ChIP = ChIP score (Experimental ChIP data) CS = Chromatin States GR = Gene reference Occup = Occupnacy locations

Value

Returns a profile plot with "Occupancy" on the y axis and DNA position on the the X- axis.

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
    }
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR", BPFrequency=DNASequenceSet)


# Computing Genome Wide
GenomeWide <- computeGenomeWideScores(DNASequenceSet = DNASequenceSet,
    genomicProfiles = GPP)

#Compute PWM Scores
PWMScores <- computePWMScore(DNASequenceSet = DNASequenceSet,
    genomicProfiles = GenomeWide,
    loci = top, chromatinState = Access)
#Compute Occupnacy
Occupancy <- computeOccupancy(genomicProfiles = PWMScores)

#Compute ChIP profiles
chipProfile <- computeChipProfile(loci = top,
    genomicProfiles = Occupancy)


#Plotting Profile
plotOccupancyProfile(predictedProfile=chipProfile,
    ChIPScore = chip,
    chromatinState = Access,
    occupancy = Occupancy,
    geneRef =geneRef)

plotOccupancyProfile(predictedProfile=chipProfile,
    ChIPScore = chip,
    chromatinState = Access,
    occupancy = Occupancy,
    geneRef = geneRef,
    colCS = c("red","blue"),
    densityGR = 60)

Heat Map of optimal Parameters

Description

plotOptimalHeatMaps will plot heat maps of optimal Parameters and highlight the optimal combination of lambdaPWM and boundMolecules

Usage

plotOptimalHeatMaps(optimalParam,contour=TRUE,col=NULL,main=NULL,layout=TRUE,overlay=FALSE)

Arguments

optimalParam

optimalParam is a list containing containing optimal matricies (or only one if only one paramter was selected). These matricies are the result of the computeOptimal function

contour

parameter is logical. Should contour lines be plotted?

col

col vector of colours to be used for each heat map. If none are specified, rainbow colours will be used. NOTE: colour vector will be recyled if not enough colours are provided.

main

main title.
layout

layout is either TRUE or FALSE specifying if stadard layout should be used or not. If TRUE, each heat map will be plotted on an individual page with a heat map scale of the right side.
overlay

overlay is either TRUE or FALSE specifying if an overlay plot should be produced. The overlay plot takes the top 10 percent of best performing parameters per scoring metric and overlays them in a single plot. The resulting plots shows the optimal set of paramters for all metrics combined.

Details

Once the optimal set of Parameters ( lambdaPWM and boundMolecules ), it is possible to plot the results in the form of a heat map. Each heat map will be plotted in a seperate page if layout = TRUE, If layout= FALSE, it is up to the user to define how they wish to layout there heat maps.

Value

Returns a heat map of optimal combinations of lambdaPWM and boundMolecules. The x axis represents the different value assigned to lambda ( lambdaPWM ) and the y axis represents the different values to boundMolecules ( boundMolecules ).

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
    if (!requireNamespace("BiocManager", quietly=TRUE))
        install.packages("BiocManager")
    BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
    }
library(BSgenome.Dmelanogaster.UCSC.dm6)
DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)

#Building data objects
GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR",BPFrequency=DNASequenceSet)

#Computing Optimal set of Parameters
optimalParam <- computeOptimal(genomicProfiles = GPP,
    DNASequenceSet = DNASequenceSet,
    ChIPScore = chip,
    chromatinState = Access,
    parameterOptions = OPP,
    parameter = "all",
    peakMethod="moving_kernel")
plotOptimalHeatMaps(optimalParam)

Accessor method for the PFM slot in a genomicProfiles object

Description

Accessor method for the PFM slot in a genomicProfiles object

Usage

PositionFrequencyMatrix(object)

Arguments

object

object is a genomicProfiles object

Details

After creating a genomicProfiles object, it is possible to access the Position Frequency Matrix slot. However this slot will be empty if the genomicProfiles object was built using directly a Position Weight Matrix. See genomicProfiles

Value

Returns the Position Frequency Matrix (PFM slot) used to compute the PositionWeightMatrix in a genomicProfiles object

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Loading data
data(ChIPanalyserData)
#Loading PFM files
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#Building genomicProfiles object
GPP<-genomicProfiles(PFM=PFM,PFMFormat="JASPAR")
# Accessing Slot
PositionFrequencyMatrix(GPP)

~~ Methods for Function PositionFrequencyMatrix ~~

Description

Accessor method for the PFM slot in a genomicProfiles object

Methods

PositionFrequencyMatrix(object)

Setter method for the PFM slot in a genomicProfiles object

Description

Setter method for the PFM slot in a genomicProfiles object

Usage

PositionFrequencyMatrix(object)<- value

Arguments

object

object is a genomicProfiles object

value

value can be of two forms. Either a matrix in the form of a Position Frequency Matrix or a path/to/file character string.

Details

The Position Frequency Matrix is one of the fundamental object that needs to be supplied to a genomicProfiles. If after building a genomicProfiles, only the Position Frequency Matrix needs to be modified then it is possible to manually update the value of this matrix using the function above. There are two options for the type of data that may be supplied to the PFM slot: a matrix in the form of a Position Frequency Matrix (matrix with four rows - one for each base pair (ACTG) and a number of columns equal to the number of sites in the binding site), or it is possible (also recommended) to provide a path to the file containing the Position Frequency Matrix. This Position Frequency Matrix file may come in multiple form such as RAW, Transfac or JASPAR. WARNING: if a genomicProfiles object has already been created and only the PFM is supplied/updated , then the Positon Weight Matrix will automatically updated as well.

Value

Returns a genomicProfiles with an updated PFM slot (as described above this will lead to an updated PositionWeightMatrix).

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

#Loading data
data(ChIPanalyserData)
#Loading PFM files
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#Building genomicProfiles object
# NOT ADVISED!!!! PLEASE PARSE PFM AND PFMFormat together
GPP<-genomicProfiles(PFMFormat = "JASPAR")
#Setting PFM
PositionFrequencyMatrix(GPP) <- PFM

~~ Methods for Function PositionFrequencyMatrix<- ~~

Description

Setter method for the PFM slot in a genomicProfiles object

Methods

PositionFrequencyMatrix(object)<-"path/to/file/"
PositionFrequencyMatrix(object)<-value

Accessor Method for the PWM slot in a genomicProfiles object

Description

Accessor Method for the PWM slot in a genomicProfiles object

Usage

PositionWeightMatrix(object)

Arguments

object

object is a genomicProfiles

Details

After creating a genomicProfiles object, it is possible to access the Position Weight Matrix stored in this slot. This slot should always contain something. This slot is either supplied by user or directly computed from a Position Frequency Matrix when supplied.

Value

Returns a matrix in the form of a Position Weight Matrix

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Loading data
data(ChIPanalyserData)
#Loading PFM files
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#Building genomicProfiles object
GPP<-genomicProfiles(PFM=PFM,PFMFormat="JASPAR")
# Accessing Slot
PositionWeightMatrix(GPP)

~~ Methods for Function PositionWeightMatrix ~~

Description

Accessor Method for the PWM slot in a genomicProfiles object

Methods

PositionWeightMatrix(object)

Setter Method for the PositionWeightMatrix slot in a genomicProfiles object

Description

Setter Method for the PositionWeightMatrix slot in a genomicProfiles object

Usage

PositionWeightMatrix(object) <- value

Arguments

object

object is a genomicProfiles object

value

value is a matrix in the form of a Position Weight Matrix.

Details

If a Position Weight Matrix is readily available, it is possible to directly assign this matrix to the PWM slot. However, this is only possible if a genomicProfiles object has already been created. In that case, we advise to first create a genomicProfiles object. It should be noted that this Position Weight Matrix will be automatically computed from a Position Frequency Matrix. If no Position Frequency Matrix are available, then a Position Weight Matrix can be directly assigned to this slot.

Value

Returns a genomicProfiles object with an updated value for the PWM slot

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Building genomicProfiles object
GPP <- genomicProfiles()
#Setting PWM to PositionWeightMatrix slot
PWM <- matrix(runif(32,-10,20), ncol=8)
rownames(PWM) <- c("A","C","T","G")
PositionWeightMatrix(GPP) <- PWM

~~ Methods for Function PositionWeightMatrix<- ~~

Description

Setter Method for the PositionWeightMatrix slot in a genomicProfiles object

Methods

PositionWeightMatrix(object)<-value

Pre-processing ChIP-seq data

Description

processingChIP will process and extract ChIP scores at a set of loci of interest.

Usage

processingChIP(profile,loci=NULL,reduce=NULL,
     peaks=NULL,chromatinState=NULL,parameterOptions=NULL,
     cores=1)

Arguments

profile

profile is a path to a UCSC format file, a GRanges or data frame. The input data should contain 4 columns:chromosome, start , end and score.

loci

loci is GRanges describing the loci at which ChIP scores should be extracted. If NULL, a set of Loci will extracted from profile. The data provided will then be split into bins of width equal to lociWidth (Default 20kbp) Default=NULL

reduce

reduce is a the top regions to select based on the mean ChIP score. If peaks are provided, regions overlappling with known peaks will be selected based on highest ChIP score. If NULL, all regions will be considered. Default=NULL

parameterOptions

parameterOptions is an parameterOptions object containing chip Parameters to be parsed for ChIP score extraction. If NULL, parameterOptions will be built internally with default ChIP extraction parameters (see chipSmooth, chipSd and chipMean) Default=NULL

peaks

peaks is a path to UCSC format file or a GRanges object containing location of ChIP peaks. Default=NULL

chromatinState

chromatinState is a GRanges containing Accessible DNA or chromatin States If provided, regions will be selected only if they contain accessible DNA. Default=NULL

cores

cores is the number of cores used to extract ChIP scores. Default = 1

Details

When using computeOptimal, it is required to supply real ChIP data in order to have a point of comparison. The corralation and MSE Scores are computed based of how well the model fits biological data. processingChIP will extract this data from ChIP data at loci of interest. When using the reduce option, this function will only select the top regions based on peak height or mean ChIP score. processingChIP will also extract maxSignal and backgroundSignal from ChIP data and parse it to an parameterOptions object.

Value

Returns a ChIPScore object containing extracted (and normalised) ChIP scores, the loci of interest and newly extracted Parameters(e.g. maxSignal)

Author(s)

Patrick C.N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Data extraction
data(ChIPanalyserData)

## Extracting ChIP scores at loci of interest

ChIP<-processingChIP(profile=chip, loci=top)

Estimating Accuracy of predicted Profiles

Description

profileAccuracyEstimate will compare the predicted ChIP-seq-like profile to real ChIP-seq data and return a set of metrics describing how accurate the predicted model is compared to real data.

Usage

profileAccuracyEstimate(genomicProfiles,ChIPScore,
      parameterOptions=NULL,method="all",cores=1)

Arguments

genomicProfiles

genomicProfiles is the result of computeChIPProfile

ChIPScore

ChIPScore is the result of processingChIP. Extracted/Normalised experimental ChIP scores.

parameterOptions

parameterOptions is a parameterOptions object for paramter specification.

method

method is the method that will be used to assess model quality agianst ChIP-seq data. Method can be one of the following: pearson, spearman, kendall, ks, geometric,fscore, MSE,or all.Fscore contains f-score, precision,recall, MCC, Accuracy and AUC ROC.
cores

cores is the number of cores used to extract ChIP scores. Default = 1

Details

In order to assess the quality of the model against experimental ChIP-seq data, ChIPanalyser offers a wide range of method to choose from. These methods are also used when computing optimal paramters.

Value

Returns list of goodness of fit metrics for each loci and each parameter selected.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

  if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
      if (!requireNamespace("BiocManager", quietly=TRUE))
          install.packages("BiocManager")
      BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
      }
  library(BSgenome.Dmelanogaster.UCSC.dm6)
  DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
  # Building genomicProfiles object
  GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR", BPFrequency=DNASequenceSet)


  # Computing Genome Wide
  GenomeWide <- computeGenomeWideScore(genomicProfiles = GPP,
      DNASequenceSet = DNASequenceSet)

  #Compute PWM Scores
  PWMScores <- computePWMScore(genomicProfiles = GenomeWide,
       DNASequenceSet = DNASequenceSet, loci = top, chromatinState = Access)
  #Compute Occupnacy
  Occupancy <- computeOccupancy(genomicProfiles = PWMScores)

  #Compute ChIP profiles
  chipProfile <- computeChIPProfile(genomicProfiles=Occupancy,loci=top)
#Estimating accuracy estimate
AccuracyEstimate <- profileAccuracyEstimate(genomicProfiles = chipProfile,
    ChIPScore = chip,
    occupancyProfileParameters = OPP)

~~ Methods for Function profiles ~~

Description

Accessor method for profiles in a genomicProfiles object

Methods

profiles(oject)

Computed PWM scores, Occupancy or ChIP-seq like profiles for loci of interest and paramter combination of interest.

Accessor Method for a PWMpseudocount slot in a parameterOptions

Description

Accessor Method for a PWMpseudocount slot in a parameterOptions

Usage

PWMpseudocount(object)

Arguments

object

object is a parameterOptions object.

Details

In the context of Position Weight Matricies, the pseudocount is used to avoid 0 probabilities during the transformation of Position Frequency Matrix to a Position Probability Matrix and finally to a Postion Weight Matrix. It is essentially a sample correction that is added in the case of small sample size. The effect of the base pair to which a pseudocount was assigned will not influence the model nor will create mathematical issues such as infinities or zero division. Default is set at 1.

Value

Returns the value assigned to a PWMpseudocount slot in a parameterOptions object

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(PWMpseudocount=0)
#Accessing slot value
PWMpseudocount(GPP)

~~ Methods for Function PWMpseudocount ~~

Description

Accessor Method for a PWMpseudocount slot in a parameterOptions

Methods

PWMpseudocount(object)

Setter Method for the pseudocount slot in a parameterOptions object

Description

Setter Method for the pseudocount slot in a parameterOptions object

Usage

PWMpseudocount(object) <- value

Arguments

object

object is a parameterOptions object

value

value is a numeric value that will be assigned to the pseudocount slot. Defualt is set at 1

Details

In the context of Position Weight Matricies, the pseudocount is used to avoid 0 probabilities during the transformation of Position Frequency Matrix to a Position Probability Matrix and finally to a Postion Weight Matrix. It is essentially a sample correction that is added in the case of small sample size. The effect of the base pair to which a pseudocount was assigned will not influence the model nor will create mathematical issues such as infinities or zero division.

Value

Returns a parameterOptions object with an updated value for the pseudocount slot.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions( PWMpseudocount=0)
#Setting Value for new PWMpseudocount
PWMpseudocount(GPP) <- 1

~~ Methods for Function PWMpseudocount<- ~~

Description

Setter Method for the pseudocount slot in a parameterOptions object

Methods

PWMpseudocount(object)<-value

Accessor method for the PWMThreshold slot in a parameterOptions object

Description

Accessor method for the PWMThreshold slot in a parameterOptions object

Usage

PWMThreshold(object)

Arguments

object

object is a parameterOptions object

Details

The computePWMScore function requires a so-called PWM Threshold. This threshold represents the Threshold at which PWM Score should be selected. The PWMThreshold is a positive numeric value (between 0 and 1. If set at 0, all sites will be selected. If set at 0.7 (Default value), then 70 % of PWM Score (and by extension binding sites) will be IGNORED. The top 30 % will be selected.

Value

Returns the value assinged to the PWMThreshold slot in a parameterOptions object

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(PWMThreshold=0.7)
#Accessing Value for  PWMThreshold
PWMThreshold(GPP)

~~ Methods for Function PWMThreshold ~~

Description

Accessor method for the PWMThreshold slot in a parameterOptions object

Methods

PWMThreshold(object)

Setter Method for the PWMThreshold slot in a parameterOptions object

Description

Setter Method for the PWMThreshold slot in a parameterOptions object

Usage

PWMThreshold(object) <- value

Arguments

object

object is a parameterOptions object

value

value is a numeric value (between 0 and 1) to be assigned to the PWMThreshold slot in parameterOptions object. Default is set at 0.7

Details

The computePWMScore function requires a so-called PWM Threshold. This threshold represents the Threshold at which PWM Score should be selected. The PWMThreshold is a positive numeric value (between 0 and 1. If set at 0, all sites will be selected. If set at 0.7 (Default value), then 70 % of PWM Score (and by extension binding sites) will be IGNORED. The top 30 % will be selected.

Value

Returns parameterOptions objetc with an updated value for the PWMThreshold slot

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(PWMThreshold=0.7)
#Setting Value for new PWMThreshold
PWMThreshold(GPP) <- 0.8

~~ Methods for Function PWMThreshold<- ~~

Description

Setter Method for the PWMThreshold slot in a parameterOptions object

Methods

PWMThreshold(object)<-value

Accessor Method for the removeBackground slot in a parameterOptions object

Description

Accessor Method for the removeBackground slot in a parameterOptions object

Usage

removeBackground(object)

Arguments

object

object is a parameterOptions object

Details

A numeric value describing a threshold at which Occupancy signals must be removed (Default is set at 0). The removal of Occupancy signals will occur when computing computeOccupancy (see computeOccupancy function)

Value

Returns the value assigned to the removeBackground slot in a parameterOptions object

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Building parameterOptions object
OPP <- parameterOptions()
#Accessing  Value for removeBackground
removeBackground(OPP)

~~ Methods for Function removeBackground ~~

Description

Accessor Method for the removeBackground slot in a parameterOptions object

Methods

removeBackground(object)

Setter Method for the removeBackground slot in a parameterOptions object

Description

Setter Method for the removeBackground slot in a parameterOptions object

Usage

removeBackground(object) <-value

Arguments

object

object is an parameterOptions object

value

value is positive numerical value to be assigned to the removeBackground slot in a parameterOptions object. Default is set a 0.

Details

A numeric value describing a threshold at which Occupancy signals must be removed (Default is set at 0). The removal of Occupancy signals will occur when computing computeOccupancy (see computeOccupancy function)

Value

Returns an parameterOptions object with an updated value for the removeBackground slot

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Building parameterOptions object
OPP <- parameterOptions()
#Setting new Value for removeBackground
removeBackground(OPP) <- 0.1

~~ Methods for Function removeBackground<- ~~

Description

Setter Method for the removeBackground slot in a parameterOptions object

Methods

removeBackground(object)<-value

Accessor Method for the scores slot in a ChIPScore object

Description

Setter Method for the scores slot in a ChIPScore object

Usage

scores(object)

Arguments

object

object is ChIPScore object

Details

When using the processingChIP, this functions will return a name list of normalised ChIP scores at loci of interest. This functions enalbles you to extract those scores from the ChIPScore object.

Value

Returns the value assigned to the scores slot in a ChIPScore object.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94.

Examples

# Loading data
data(ChIPanalyserData)

chip<-processingChIP(chip,top)
str(scores(chip))

~~ Methods for Function scores ~~

Description

Accessor method for scores slot in a ChIPScore object.

Methods

scores(object)

Extracted and normalised ChIP scores at loci of interest.

Searching function for Sites above threshold and predicted ChIP-seq Profiles

Description

searchSites is function enabling quick extraction and search for parameter combinations and/or loci in any genomicProfiles object from computeOccupancy onwards.

Usage

searchSites(Sites,lambdaPWM="all",BoundMolecules="all", Locus="all")

Arguments

Sites

Sites is either a genomicProfiles or the result of computeOptimal

lambdaPWM

lambdaPWM is a numeric vector describing the ScalingFactors that should be searched within Sites.

BoundMolecules

BoundMolecules is a numeric vector describing the BoundMolecules that should be searched within Sites.

Locus

Locus is a character vector describing the Loci that should be searched within Sites.

Details

When testing numerous combinations of lambdaPWM and boundMolecules on top of many loci, it can become challenging to navigate the large data output searchSites will make searching in this slot a lot easier. If all arguments are left at their default value of "all", then all Parameters will be searched thus returning the full list of Sites above threshold. If a value for lambdaPWM is user provided then only this lambdaPWM will be selected (all boundMolecules and loci will also be selected). searchSites also works on the result of computeOptimal.

Value

Returns object of same time as parsed to this function with only the parameters and/or loci selected.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

#Data extraction
data(ChIPanalyserData)
# path to Position Frequency Matrix
PFM <- file.path(system.file("extdata",package="ChIPanalyser"),"BEAF-32.pfm")
#As an example of genome, this example will run on the Drosophila genome

  if(!require("BSgenome.Dmelanogaster.UCSC.dm6", character.only = TRUE)){
      if (!requireNamespace("BiocManager", quietly=TRUE))
          install.packages("BiocManager")
      BiocManager::install("BSgenome.Dmelanogaster.UCSC.dm6")
      }
  library(BSgenome.Dmelanogaster.UCSC.dm6)
  DNASequenceSet <- getSeq(BSgenome.Dmelanogaster.UCSC.dm6)
  # Building genomicProfiles object
  GPP <- genomicProfiles(PFM=PFM,PFMFormat="JASPAR", BPFrequency=DNASequenceSet)


  # Computing Genome Wide
  GenomeWide <- computeGenomeWideScore(genomicProfiles = GPP,
      DNASequenceSet = DNASequenceSet)

  #Compute PWM Scores
  PWMScores <- computePWMScore(genomicProfiles = GenomeWide,
       DNASequenceSet = DNASequenceSet, loci = top, chromatinState = Access)
  #Compute Occupnacy
  Occupancy <- computeOccupancy(genomicProfiles = PWMScores)
  searchSites(Occupancy,ScalingFactor=c(1,4), BoundMolecules = c(1,100),
      Locus="eve")

  #Compute ChIP profiles
  chipProfile <- computeChIPProfile(genomicProfiles=Occupancy,loci=top)
  searchSites(chipProfile,ScalingFactor=c(1,4), BoundMolecules = c(1,100),
      Locus="eve")

optimalParam <- computeOptimal(genomicProfiles = GPP,
        DNASequenceSet = DNASequenceSet,
        ChIPScore = chip,
        chromatinState = Access,
        parameterOptions = OPP,
        parameter = "all",
        peakMethod="moving_kernel")

searchSites(optimalParam,ScalingFactor=c(1,4), BoundMolecules = c(1,100),
    Locus="eve")

setChromatinStates

Description

setChromatinStates sets chromatin state affinity values to a GRanges object.

Usage

setChromatinStates(population,chromatinStates)

Arguments

population

Population list containing all individuals and associated parameter. Must contain chromatin state affinity values. See generateStartingPopulation.

chromatinStates

GRanges object containing chromatin state locations.

Details

Chromatin states can be loaded into R as a GRanges object. Each range represents the extent of a certain chromatin state and the chromatin state type should be assigned to a meta data column called "name". The affinity values names should be set accordingly.

Value

Returns a GRange object with affinity scores for each chromatin state range. Affinity scores are placed in the DNAAffinity meta data column.

Author(s)

Patrick C.N. Martin

Examples

library(ChIPanalyser)
# Input data 
data(ChIPanalyserData)

pop <- 10
params <- c("N","lambda","PWMThreshold", paste0("CS",seq(1:11)))
start_pop <- generateStartingPopulation(pop, params)

cs <- setChromatinStates(start_pop,cs)

~~ Methods for Function show ~~

Description

Show methods for various objects

Methods

signature(object = "ChIPScore")
signature(object = "genomicProfiles")
signature(object = "parameterOptions")

singleRun

Description

singleRun runs ChIPanalyser after optimal paramters have been found by the evolve function.

Usage

singleRun(indiv,DNAAffinity,
            genomicProfiles,DNASequenceSet,
            ChIPScore,fitness="all")

Arguments

indiv

Population list containing the top scoring individual. Note that this should be a list of length 1 containing another list with all parameter values.

DNAAffinity

GRanges object as outputed by the setchromatinStates.

genomicProfiles

genomicProfiles object containing PWM scores and other desired metrics. Note that PWMThreshold, lambda and N will be overwritten using values from indiv.

DNASequenceSet

DNA string set object containing DNA sequence of interest.

ChIPScore

ChIPScore object as outputed by the processingChIP function.

fitness

character string describing which metric should be used to assess fitness and should be one of the following:"geometric","ks","MSE","pearson","spearman","kendall", "recall","precesion","fscore","MCC","Accuracy" or "AUC".

Details

Once the genetic algorithm has been optimised, the top individual may be run on its own to get predicted ChIP profiles. The use of this function requires a few extract steps in order to predict ChIP profiles.

First, the index of the top individual should be extracted (see getHighestFitnessSolutions). Second, using this index, subset top individual from GA population. Note this should be done using "[]" single bracket notation as, a list of length 1 containing another list with all parameter values is required for the next steps. Yes, this is might seem annoying but the functions were design for list structures... Third, setchromatinStates using the top individual list. This will add chromatin affinity values to your chromatinState GRanges. Use this new chromatinState object as your new chromatinState object. Fourth, parse your indiv list object to singleRun.

Value

Return a list with three elements. First element contains a genomicProfiles object with occupancy scores. Second element contains a genomicProfiles objecy with ChIP profile scores. Third element contains a goodness of fit metrics.

Author(s)

Patrick C.N. Martin <[email protected]

Examples

library(ChIPanalyser)
data(ChIPanalyserData)
# See GA vignette for usage

Get Training and Testing data from ChIPscore objects

Description

splitData splits processed ChIP data into training and testing sets.

Usage

splitData(ChIPscore,dist = c(80,20), as.proportion = TRUE)

Arguments

ChIPscore

ChIPscore object as returned by processingChIP

dist

If as.proportion is to TRUE, split the data into desired proportions. Default sets 80% training and 20% testing. If as.proportion is to FALSE,a vector of 4 numeric values describing start and end of training and testing respectively.

as.proportion

Logical describing if values provided to dist should be treated as % of training and testing or if dist should be considered as start and end of loci selected for training and testing respectively.

Value

Returns a named list of ChIPScore objects

* trainingSet = ChIPscore containing training set * testingSet = ChIPscore containing testing set.

Author(s)

Patrick C.N. Martin <[email protected]

Examples

library(ChIPanalyser)
data(ChIPanalyserData)
# See GA vignette for usage 
test <- processingChIP(chip,top)
usingDist <- splitData(test, dist = c(50,50),as.proportion = TRUE )
usingIndex <- splitData(test, dist = c(1,2,3,4),as.proportion = FALSE )

Accessor method of the stepSize slot in parameterOptions object

Description

Accessor method of the stepSize slot in parameterOptions object

Usage

stepSize(object)

Arguments

object

object is a parameterOptions object.

Details

It possible to restrict the size of the ChIP-seq-like profile produced by computeChIPProfile. Instead of returning ChIP-seq like score for each base pair, it is possible to skip base pairs and only return the predicted enrichement score for every "n" base pair (n is the value assigned to stepSize). This will reduce the size of the output data (unless step size is very large, this will not affect the accuracy of the model). Default is set at 10 base pairs.

Value

Returns the value assigned to the stepSize slot in a parameterOptions

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Building parameterOptions  object
OPP <- parameterOptions()
#Setting new Value for maxSignal
stepSize(OPP)

~~ Methods for Function stepSize ~~

Description

Accessor method of the stepSize slot in parameterOptions object

Methods

stepSize(object)

Setter Method for the stepSize slot in a parameterOptions

Description

Setter Method for the stepSize slot in a parameterOptions

Usage

stepSize(object) <- value

Arguments

object

object is a parameterOptions object

value

value is a positive numeric value that will be assigned to the stepSize slot in a parameterOptions object. Default is set at 10 base pairs.

Details

It possible to restrict the size of the ChIP-seq-like profile produced by computeChIPProfile. Instead of returning ChIP-seq like score for each base pair, it is possible to skip base pairs and only return the predicted enrichement score for every "n" base pair (n is the value assigned to stepSize). This will reduce the size of the output data (unless step size is very large, this will not affect the accuracy of the model). Default is set at 10 base pairs.

Value

Returns a parameterOptions object with an updated value for the stepSize slot.

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Building parameterOptions  object
OPP <- parameterOptions()
#Setting new Value for maxSignal
stepSize(OPP) <- 20

~~ Methods for Function stepSize<- ~~

Description

Setter Method for the stepSize slot in a parameterOptions

Methods

stepSize(object)<-value

Accessor Method for the strandRule slot in a parameterOptions object

Description

Accessor Method for the strandRule slot in a parameterOptions object

Usage

strandRule(object)

Arguments

object

object is a parameterOptions object

Details

When computing the PWM Scores and if whichstrand is set to "+-", strandRule will determine how to handle both strands ( one of three options : "mean", "max", "sum"). If set to "mean", the average PWM Score of both strand will be computed. If set to "max", the highest PWM score between each strand will be selected and finally "sum" will sum both score together. Default set at "max"

Value

Returns the value assigned to strandRule slot (one of three options : "mean", "max", "sum") in a parameterOptions object

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions( strandRule="max")
#Accesssing Value for strandRule
strandRule(GPP)

~~ Methods for Function strandRule ~~

Description

Accessor Method for the strandRule slot in a parameterOptions object

Methods

strandRule(object)

Setter method for the strandRule slot in a parameterOptions object.

Description

Setter method for the strandRule slot in a parameterOptions object.

Usage

strandRule(object) <- value

Arguments

object

object is a parameterOptions object

value

value is a character string and can be one of the following ‘mean’, ‘max’, ‘sum’. This will only apply if whichstrand is ‘+-’. Default set at ‘max’

Details

When computing the PWM Scores and if whichstrand is set to ‘+-’, strandRule will determine how to handle both strands ( one of three options : ‘mean’, ‘max’, ‘sum’). If set to ‘mean’, the average PWM Score of both strand will be computed. If set to ‘max’, the highest PWM score between each strand will be selected and finally ‘sum’ will sum both score together. Default set at ‘max’

Value

Returns a parameterOptions object with an updated value for the strandRule slot

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions(strandRule="max")
#Setting New Value for strandRule
strandRule(GPP) <- "mean"

~~ Methods for Function strandRule<- ~~

Description

Setter method for the strandRule slot in a parameterOptions object.

Methods

strandRule(object)<-value

Accessor method for the whichstrand slot in a parameterOptions object

Description

Accessor method for the whichstrand slot in a parameterOptions object

Usage

whichstrand(object)

Arguments

object

object is a parameterOptions object

Details

PWM Score may be computed on either the positive strand ("+"), the negative strand ("-") or on both strands ("+-").

Value

Returns on which strand PWM Scores should be computed ( whichstrand in a parameterOptions object)

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions( whichstrand="+-")
#Setting New Value for whichstrand
whichstrand(GPP)

~~ Methods for Function whichstrand ~~

Description

Accessor method for the whichstrand slot in a parameterOptions object

Methods

whichstrand(object)

Setter method for the whichstrand slot in a parameterOptions object

Description

Setter method for the whichstrand slot in a parameterOptions object

Usage

whichstrand(object) <- value

Arguments

object

object is a parameterOptions object

value

value is a character string specifying which strand should be used to compute PWM Scores. The three available options are the following: "+","-" or "+-". Default is "+-"

Details

PWM Score may be computed on either the positive strand ("+"), the negative strand ("-") or on both strands ("+-").

Value

Returns a parameterOptions object with an updated value for the whichstrand slot

Author(s)

Patrick C. N. Martin <[email protected]>

References

Zabet NR, Adryan B (2015) Estimating binding properties of transcription factors from genome-wide binding profiles. Nucleic Acids Res., 43, 84–94. Patrick C.N. Martin and Nicolae Radu Zabe (2020) Dissecting the binding mechanisms of transcription factors to DNA using a statistical thermodynamics framework. CSBJ, 18, 3590-3605.

Examples

# Loading data
data(ChIPanalyserData)

#Building data objects
GPP <- parameterOptions( whichstrand="+-")
#Setting New Value for whichstrand
whichstrand(GPP) <- "+"

~~ Methods for Function whichstrand<- ~~

Description

Setter method for the whichstrand slot in a parameterOptions object

Methods

whichstrand(object)<-value