callPeaks: Call peaks using transformed, background corrected, and...

In jianhong/NADfinder: Call wide peaks for sequencing data

Call peaks using transformed, background corrected, and smoothed ratios with biological replicates

Description

Use limma to calculate p-values for NADs

Usage

callPeaks(
  se,
  backgroundCorrectedAssay = "bcRatio",
  normalization.method = "quantile",
  N = 100,
  cutoffAdjPvalue = 1e-04,
  countFilter = 1000,
  combineP.method = "minimump",
  smooth.method = "loess",
  lfc = log2(1.5),
  ...
)

Arguments

se	An object of RangedSummarizedExperiment with assays of raw counts, tranformed ratios, background corrected ratios, smoothed ratios and z-scores. It should be an element of output of smoothRatiosByChromosome
backgroundCorrectedAssay	character(1). Assays names for background corrected log2-transformed ratios, CPMRatios or OddRatios.
normalization.method	character(1) specifying the normalization method to be used. Choices are "none", "scale", "quantile" or "cyclicloess". See normalizeBetweenArrays for details.
N	numeric(1) or integer(1). The number of neighboring windows used for loess smoothing or the inverse of the critical frequencies of the low pass filter for butterworth filter. 1/N is a cutoff at 1/N-th of the Nyquist frequency. Default 100.
cutoffAdjPvalue	numeric(1). Cutoff adjust p-value.
countFilter	numeric(1). Cutoff value for mean of raw reads count in each window.
combineP.method	A method used to combine P-values. Default minimump
smooth.method	A method used to smooth the ratios. Choices are "loess", "none" and "butterworthfilter".
lfc	the minimum log2-fold-change that is considered scientifically meaningful
...	Parameter not used.

Details

By default, use the mean smoothed ratio for each peak region to calculate p-values

Value

An object of GRanges of peak list with metadata "AveSig", "P.Value", and "adj.P.Val", where "AveSig" means average signal such as average log2OddsRatio, log2CPMRatio or log2Ratio.

Author(s)

Jianhong Ou, Haibo Liu and Julie Zhu

Examples

data(triplicate.count)
se <- triplicate.count
se <- log2se(se, transformation = "log2CPMRatio",
             nucleolusCols = c("N18.subsampled.srt-2.bam",
             "N18.subsampled.srt-3.bam",
             "N18.subsampled.srt.bam"),
             genomeCols = c("G18.subsampled.srt-2.bam",
             "G18.subsampled.srt-3.bam",
             "G18.subsampled.srt.bam"))
se<- smoothRatiosByChromosome(se, chr="chr18")
#add some variability to the data since the triplicate.count data was created using one sample only
assays(se[[1]])$bcRatio[,2] <- assays(se[[1]])$bcRatio[,2] + 0.3
assays(se[[1]])$bcRatio[,3] <- assays(se[[1]])$bcRatio[,3] - 0.3
peaks <- callPeaks(se[[1]],
                cutoffAdjPvalue=0.001, countFilter=10)

jianhong/NADfinder documentation built on Nov. 2, 2024, 12:09 a.m.