mulitpleTestingCorrections.plotSignifFeatures: Plot number of significant features as a function of the...

In jvanheld/stats4bioinfo: Utilities for the book "Statistics for bioinformatics"

Description

Plot the number of significant features as a function of the control criterion (nominal p-value, e-value, fdr, ...).

Usage

mulitpleTestingCorrections.plotSignifFeatures(multitest.result,
  main = "Significant features", xlab = "Significance threshold",
  ylab = "Significant features", alpha = 0.05, plot.legend = TRUE,
  legend.corner = "topleft", legend.cex = 1, xlim = NULL,
  plot.pch = c(p.value = 2, fdr = 4, qval.0 = 3, e.value = 1, fwer = 20),
  plot.col = c(p.value = "#000000", fdr = "#888888", qval.0 = "#666666",
  e.value = "#BBBBBB", fwer = "#444444"), plot.elements = c("p.value", "fdr",
  "qval.0", "e.value", "fwer"), ...)

Arguments

multitest.result	the list returned by the function multipleTestingCorrections().
...	Additional parameters are passed to plot()
main="Significant	features" main title of the plot
xlab="P-value	derived statistics"
ylab="Significant	features"
alpha=0.05	Threshold of significance (alpha).
plot.legend=TRUE	Plot a legend indicating the number of features declared significant with the alpha threshold on the selected statistics.
legend.corner="topleft"	corner wher the legend has to be placed.
legend.cex=1	Font size for the legend.
plot.pch=c(p.value=2, e.value=1, fwer=20, fdr=4, qval.0=3)	Specific characters to distinguish the plotted statistics.
plot.col=c(p.value='#000000', e.value='#BBBBBB', fwer='#444444', fdr='#888888', qval.0='#666666')	Specific colors or gray levels to distinguish the plotted statistics.
plot.elements=c("p.value", "fdr", "qval.0", "e.value", "fwer")	Selection of elements to display on the plot.

Author(s)

Jacques van Helden ([e-mail])

Examples

## To obtain the input list (multitest.result), see the documentatio of
example(multipleTestingCorrections)

## Plot all the multiple testing corrections at once
mulitpleTestingCorrections.plotSignifFeatures(multitest.result)

## Compare e-value and FWER
mulitpleTestingCorrections.plotSignifFeatures(multitest.result, plot.elements=c("e.value","fwer"))

## Compare e-value and FDR
mulitpleTestingCorrections.plotSignifFeatures(multitest.result, plot.elements=c("fdr","e.value"))

## Compare Benjamini-Hochberg (qval.0) and Storey-Tibshirani (fdr) estimates of FDR
mulitpleTestingCorrections.plotSignifFeatures(multitest.result, plot.elements=c("fdr","qval.0"))

jvanheld/stats4bioinfo documentation built on May 20, 2019, 5:16 a.m.