Object Oriented Microarray Library: Rank Coherence

The slide-comparison2 module provides the definition of the slide.conparison2 class. See the bottom of the page for an example of how the class can be used.

Attributes

data

A list containing x and y components, representing smooth t-scores.

up

A vector, representing the rank coherence values for the agreement between two analyses.

random

A vector, representing the rank coherence values of simulated independently ranked data.

name1

A character string; the name of the first experiment set.

name2

A character string; the name of the second experiment set.

Methods

slide.comparison2(one, two, filt)

The constructor takes two arguments of a class (such as the replicate.ratio class) derived from two.groups. The optional filt argument allows you to remove certain genes from the comparison of rankings. The constructor computes the rank coherence between the two analysis objects, and alos simulates random agreement for compariosn purposes. (This is probabnly no longer a smart thing to do.)

plot(object, tag, ccl, vlines, hlines, ...)

The plot method produces two graphs of the object. The first is a direct comparison of the two sets of t-statistics; the second is the rank coherence plot. The optional tag argument is prepended to the names of the analysis objects for use in the title of the latter graph. The cll argument seems to no longer be used for anything; maybe this will change in the future. The optional hlines and vlines arguments are passed to abline to add horizontal or vertical lines to the graph, respectively.

image(object, tag, ...)

The image method is a misnomer; it just plots the rank coherence plot without any bells and whistles.

Functions

ratio.of.ratios(cs1, cs2, name)

Thisn is a specialized function used to extract values from an experiment using a common reference channel. Instead of directly computing log(A/B), we must compute log((A/R1) / (B/R2)) where R1 and R2 are the two reference channels which should nominally cancel out. Both cs1 and cs2 should be complete.slide objects. The return value is a two.groups object.

Description

An object of the slide.comparison2 class encapsulates the reuslts of a rank coherence analysis comparing two experiments that are both ranking the same sets of genes with smooth t-statistics.