DpeakFit-class: Class "DpeakFit"

In dpeak: dPeak (Deconvolution of Peaks in ChIP-seq Analysis)

Description

This class represents deconvolution model fit.

Objects from the Class

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

Slots

fits:

Object of class "list", representing list of deconvolution fits of all possible models for each peak.

optFit:

Object of class "list", representing list of fits of the optimal model for each peak.

optMu:

Object of class "list", representing list of binding sites of the optimal model for each peak.

optPi:

Object of class "list", representing list of relative strengths of the optimal model for each peak.

optPi0:

Object of class "list", representing list of background strengths of the optimal model for each peak.

optGamma:

Object of class "list", representing list of background proportion of the optimal model for each peak.

optDelta:

Object of class "list", representing list of read shift of the optimal model for each peak when PET=FALSE.

optSigma:

Object of class "list", representing list of read standard deviation of the optimal model for each peak when PET=FALSE.

bicVec:

Object of class "list", representing list of BIC values for each peak.

aicVec:

Object of class "list", representing list of AIC values for each peak.

fragSet:

Object of class "list", representing list of fragments for each peak.

PET:

Object of class "logical", representing whether it is paired-end tag (PET) or single-end tag (SET) data.

fragLenTable:

Object of class "table", representing distribution of fragment length when PET=TRUE.

aveFragLen:

Object of class "numeric", representing average fragment length when PET=FALSE.

Fratio:

Object of class "numeric", representing proportion of forward reads when PET=FALSE.

stackedFragment:

Object of class "list", representing number of fragments aligning to each genomic position.

peakChr:

Object of class "character", representing a vector of chromosome of each peak.

peakStart:

Object of class "numeric", representing a vector of start position of each peak.

peakEnd:

Object of class "numeric", representing a vector of end position of each peak.

estDeltaSigma:

Object of class "character", representing the approach to estimate delta and sigma parameters for SET data.

nTop:

Object of class "numeric", representing the number of candidate regions used to estimate common delta and sigma estimates.

lbDelta:

Object of class "numeric", representing a lower bound for the delta parameter.

lbSigma:

Object of class "numeric", representing a lower bound for the sigma parameter.

psize:

Object of class "numeric", representing approximate size of the binding protein of interest.

maxComp:

Object of class "numeric", representing maximum possible number of binding events in each peak region.

pConst:

Object of class "numeric", representing value to determine the plateau in the BIC curve.

iterInit:

Object of class "numeric", representing iteration number for initial estimation of binding sites.

iterMain:

Object of class "numeric", representing iteration number for main estimation of binding sites.

epsilon:

Object of class "numeric", representing criterion to stop iteration for binding site estimation.

Methods

exportPlot

signature(x = "DpeakFit", y = "missing", filename=NULL, plotType="fit", strand=FALSE, extension=1, smoothing=FALSE, threshold=0.10, nsimul=10000, seed=12345, nCore=8): draw plots of deconvolution results if plotType="fit", goodness of fit (GOF) plots if plotType="GOF", or plots of Bayesian information criterion (BIC) and Akaike information criterion (AIC) curves if plotType="BIC". Plots are exported to a PDF file (its file name is specified in filename). In deconvolution result plots, binding sites with strength larger than threshold are drawn in dark blue and other binding sites are drawn in light blue. When plotType="fit", options strand, extension, and smoothing are supported for SET data. If strand=TRUE, reads are plotted in a strand-specific manner, where reads are extended to extension from its 5' end. If smoothing=TRUE, a smoothed plot (using the smoothing spline) is provided. If strand=FALSE, strand information is ignored. For the GOF plots, nsimul fragments are simulated from the fitted model (seed indicates random seed; nCore CPUs are used for parallel computing).

show

signature(object = "DpeakFit"): provide brief summary of the object.

Author(s)

Dongjun Chung

See Also

dpeakFit, exportPeakList.

Examples

data(exampleData)
exampleFit <- dpeakFit(exampleData, maxComp = 5)

dpeak documentation built on Nov. 8, 2020, 7:45 p.m.