plot.varTestBioCond: Plot a `varTestBioCond` Object

Description

Given a varTestBioCond object, which records the results of calling hypervariable and invariant genomic intervals across ChIP-seq samples of a bioCond object, this method creates a scatter plot of observed (mean, log10(variance)) pairs from all genomic intervals, marking specifically the ones that have a significantly large or small variance. Besides, the mean-variance curve associated with the bioCond is also added to the plot, serving as a baseline to which each observed variance could be compared.

Usage

# S3 method for varTestBioCond
plot(
  x,
  padj = NULL,
  pval = NULL,
  col = alpha(c("black", "red"), 0.04),
  pch = 20,
  xlab = "Mean",
  ylab = "log10(Var)",
  args.legend = list(x = "bottomleft"),
  args.lines = list(col = "green3", lwd = 2),
  ...
)

Value

The function returns NULL.

Arguments

x: An object of class "varTestBioCond", typically a returned value from varTestBioCond.
padj, pval: Cutoff of adjusted/raw p-value for selecting significant intervals. Only one of the two arguments is effectively used; pval is ignored if padj is specified. The default is equivalent to setting padj to 0.1.
col, pch: Optional length-2 vectors specifying the colors and point characters of non-significant and significant intervals, respectively. Elements are recycled if necessary.
xlab, ylab: Labels for the X and Y axes.
args.legend: Further arguments to be passed to legend.
args.lines: Further arguments to be passed to lines.
...: Further arguments to be passed to plot.

Details

Those genomic intervals considered to be significant are actually the ones that significantly deviate from the mean-variance curve in the plot. See varTestBioCond for technical details of the associated hypothesis testing.

Examples

Run this code

data(H3K27Ac, package = "MAnorm2")
attr(H3K27Ac, "metaInfo")

## Call hypervariable and invariant genomic intervals across biological
## replicates of the GM12891 cell line.
# \donttest{
# Perform MA normalization and construct a bioCond to represent GM12891.
norm <- normalize(H3K27Ac, 5:6, 10:11)
GM12891 <- bioCond(norm[5:6], norm[10:11], name = "GM12891")

# Fit a mean-variance curve for GM12891 using the parametric method.
GM12891 <- fitMeanVarCurve(list(GM12891), method = "parametric",
                           occupy.only = TRUE)[[1]]
summary(GM12891)
plotMeanVarCurve(list(GM12891), subset = "occupied")

# Assess the observed variances of ChIP-seq signal intensities in GM12891.
res <- varTestBioCond(GM12891)
head(res)

# Inspect only the test results of occupied genomic intervals.
res <- res[GM12891$occupancy, ]
res$padj <- p.adjust(res$pval, method = "BH")
plot(res, col = scales::alpha(c("black", "red"), c(0.04, 0.5)))
# }