OCshow: Show one ore several OC curves graphically

In OCplus: Operating characteristics plus sample size and local fdr for microarray experiments

Description

Plots empirical OC curves for one or several data sets simultaneously, showing the local or global false discovery rate among the top regulated genes. This is the preferred way of comparing the OC of different analyses.

Usage

OCshow(x, ..., global = TRUE, percentage = TRUE, top = 0.1, legend, lty, col, main, xlab, ylab)

Arguments

x, ...	one or several objects created by either EOC, fdr1d, or fdr2d
global	logical value indicating whether to show the global or the local false discovery rates; note that if any of the objects to be plotted was created by EOC, only global Fdr is available.
percentage	logical switch indicating whether to show the percentage of top regulated genes or the actual numbers; note that the cutoff top is always a percentage
top	a value between 0 and 1 specifying the percentage of top-regulated genes that is to be shown in the plot
legend	a character vector giving names for each of the objects to be plotted for a legend in the left upper corner
lty, col	line styles and colors for the different OC curves
main	a plot title
xlab, ylab	axis labels

Details

Each object generated by EOC, fdr1d, and fdr2d lists for each gene a t-statistic and either a local or a global false discovery rate. The OC curves are constrcuted by ordering the genes according to the false discovery rates, and counting how many fall under a given threshold. These counts are plotted (either directly or as percentage of all genes) on the horzontal axis, while the thresholds are plotted on the vertical axis. Where appropriate, local false discovery rates are converted to global rates by simple averaging.

Author(s)

A. Ploner

See Also

EOC, fdr1d, fdr2d

Examples

# We simulate a small example with 5 percent regulated genes and
# a rather large effect size
set.seed(2003)
xdat = matrix(rnorm(50000), nrow=1000)
xdat[1:25, 1:25] = xdat[1:25, 1:25] - 2
xdat[26:50, 1:25] = xdat[26:50, 1:25] + 2
grp = rep(c("Sample A","Sample B"), c(25,25))

# Compute the different false discovery rates
# p0 is fixed
global = EOC(xdat, grp, plot=FALSE, p0=0.95)
local1d = fdr1d(xdat, grp, p0=0.95)
local2d = fdr2d(xdat, grp, p0=0.95)

# Some possible arrangements
leg = c("global","local1d","local2d")
par(mfrow=c(2,2))
OCshow(global, local1d, local2d, legend=leg, main="Default")
OCshow(global, local1d, local2d, legend=leg, percentage=FALSE, 
       main="Number of genes")
OCshow(global, local1d, local2d, legend=leg, top=1, main="All genes")
OCshow(local1d, local2d, legend=leg[2:3], global=FALSE, main="Local fdr")

OCplus documentation built on Nov. 8, 2020, 5:20 p.m.