normalize_coverage_matrix: normalize_coverage_matrix

In skummerf/GenomicWidgets: interactive visualization of genomics data

Description

Normalizes coverage matrices using one of several methods.

Usage

## S4 method for signature 'list'
normalize_coverage_matrix(mats, method = c("localRms",
  "localMean", "localNonZeroMean", "PercentileMax", "scalar", "none"),
  pct = 0.95, scalar = NULL, digits = 3)

## S4 method for signature 'matrix'
normalize_coverage_matrix(mats, method = c("localRms",
  "localMean", "localNonZeroMean", "PercentileMax", "scalar", "none"),
  pct = 0.95, scalar = NULL, digits = 3)

## S4 method for signature 'SummarizedExperiment'
normalize_coverage_matrix(mats, ...)

Arguments

mats	matrix, list of matrix, or SummarizedExperiment
method	normalization method option, see Details
pct	Percentile, only used if PercentileMax is method
scalar	vector of scalars used for normalizing each mat, only used if scalar is method
digits	number of significant digits of result to keep.
...	additional arguments to normalize_coverage_matrix

Details

Normalization choices are "localRms", "localMean", "localNonZeroMean", "PercentileMax", "scalar", and "none". localRMS will divide each row by the root mean squared values of that row. localMean will divide each row by the mean of that row. localNonZeroMean will divide each row by nonzero values in that row. PercentileMax will divide values based on percentile (given by pct argument) of the entire matrix. scalar will divide entire matrix by a scalar, given by scalar argument. This scalar could for example be a measure of the sequencing depth.

Value

Should return data in the same format as input, but now with values normalized according to the method chosen.

Author(s)

Alicia Schep

Examples

## First we'll make some coverage matrices

library(GenomicRanges)
# First read in some sample data
genomation_dir <- system.file("extdata", package = "genomationData")

samp.file <- file.path(genomation_dir,'SamplesInfo.txt')
samp.info <- read.table(samp.file, header=TRUE, sep='\t', 
                       stringsAsFactors = FALSE)
samp.info$fileName <- file.path(genomation_dir, samp.info$fileName)

ctcf.peaks = genomation::readBroadPeak(system.file("extdata",
               "wgEncodeBroadHistoneH1hescCtcfStdPk.broadPeak.gz",
               package = "genomationData"))
ctcf.peaks = ctcf.peaks[seqnames(ctcf.peaks) == "chr21"]
ctcf.peaks = ctcf.peaks[order(-ctcf.peaks$signalValue)]
ctcf.peaks = resize(ctcf.peaks, width = 1000, fix = "center")

# Make the coverage matrices
mats <- make_coverage_matrix(samp.info$fileName[1:3], ctcf.peaks, 
                     up = 500, down = 500, binsize = 25)
                     
# Now normalize:
norm_mats <- normalize_coverage_matrix(mats)                    
                     

skummerf/GenomicWidgets documentation built on May 31, 2019, 6:16 p.m.