coverage.correlation: Coverage correlation plot
In hummelma/TEQC: Quality control for target capture experiments

Description Usage Arguments Details Value Author(s) See Also Examples

Visualization of target coverage correlations between pairs of samples.

1 2	coverage.correlation(coveragelist, normalized = TRUE, plotfrac = 0.001, seed = 123, labels, main, pch = ".", cex.labels, cex.pch = 2, cex.main = 1.2, cex.corr, font.labels = 1, font.main = 2, ...)

`coveragelist`	List where each element is the output of function `coverage.target`, where option `perBase` had to be set to `TRUE`.
`normalized`	if `TRUE`, correlation of normalized target coverages will be shown; original coverages otherwise
`plotfrac`	numeric value between 0 and 1. Coverages for a fraction of `plotfrac` of all target bases are shown.
`seed`	seed for random selection of `plotfrac` bases
`labels`	sample names that are written in the diagonal panels; if missing, names of `coveragelist` are taken; if those are `NULL`, "sample 'i'" is shown
`main`	main title
`pch`	plot symbol for the scatter plots
`cex.labels, cex.pch, cex.main`	sizes of sample labels, plot symbols, main title
`cex.corr`	size of the correlation values; if missing, sizes are made proportionally to the values of (positive) correlation.
`font.labels, font.main`	fonts for sample labels and main title
`...`	further graphical parameters, e.g. limits and symbol color for the scatter plots

If normalized = TRUE, the function calculates normalized coverages: per-base coverages divided by average coverage over all targeted bases. Normalized coverages are not dependent on the absolute quantity of reads and are hence better comparable between different samples or even different experiments.

'pairs'-style plot where upper panels show scatter plot of (a randomly chosen fraction of) coverage values for pairs of samples. The lower panels show the respective Pearson correlation coefficients, calculated using all coverage values (even if not all of them are shown in the scatter plot).

Manuela Hummel m.hummel@dkfz.de

coverage.target, covered.k, coverage.hist, coverage.density, coverage.uniformity, coverage.plot

## get reads and targets
exptPath <- system.file("extdata", package="TEQC")
readsfile <- file.path(exptPath, "ExampleSet_Reads.bed")
reads <- get.reads(readsfile, idcol=4, skip=0)
targetsfile <- file.path(exptPath, "ExampleSet_Targets.bed")
targets <- get.targets(targetsfile, skip=0)

## calculate per-base coverages
Coverage <- coverage.target(reads, targets, perBase=TRUE)

## simulate another sample
r <- sample(length(reads), 0.1 * length(reads))
reads2 <- reads[-r,,drop=TRUE]
Coverage2 <- coverage.target(reads2, targets, perBase=TRUE)

## coverage uniformity plot
covlist <- list(Coverage, Coverage2)
coverage.correlation(covlist, plotfrac=0.1)