concub-class {geecc} | R Documentation |
"concub"
An object of type concub
Specifying the background population is crucial for the tests for association between factors. Usually the population is the set of all probe sets represented on a micro-array or the set of all genes in a genome. If an expression set is passed with the population-parameter, all probe sets beginning with the pattern “AFFX” (Affymetrix quality control) are removed.
Objects can be created by calls of the form new("concub", ...)
.
categories
:A named list of named lists. Each item of the outer list represents the two or three categories. Each item of the inner lists represents a variable of the category.
population
: A character vector containig the background population. As an alternative, an object with class ‘eSet’, ‘ExpressionSet’, or ‘DGEList’; background population is then set to the outcome of rownames(population)
.
keep.empty.vars
: A boolean list with names being names of categories
.
options
:Additional options for individual categories. See Details.
approx
:specifies the minimum expected value when an exact hypergeometric test (below) or the chi-squared approximation should be used. Defaults to 0.
null.model
:A formula specifying the null-model of the test.
test.result
: A list to store test results. Filled up after runConCub
.
test.result.filter
: A list to store filtered test results. Filled up after filterConCub
.
test.result.filter.heatmap
: A list to store heatmaps for further manipulation. Filled up after plotConCub
.
The last three slots are not set by the user.
signature(object = "concub")
:
creates a table containing the results of the tests.
Usage: getTable(object, na.rm=TRUE, dontshow=list())
Arguments:
an object of type concub
logical. If TRUE
(the default), rows with NA
P-values (or odds ratios) are removed
The resulting table is a data frame with 8 or 10 columns, depending on if a two- or three-way test was applied
'factor1':this column has the name of the first category
'factor2':this column has the name of the second category
'factor3':this column has the name of the third category
n.'factor1':number of items in variable of first category
n.'factor2':number of items in variable of second category
n.'factor3':number of items in variable of third category
p.value:(probably adjusted) p-value
log2.odds.ratio:log2 of the sample odds ratio
n.tags:number of items at position x_{1,1,1} or x_{1,1}
tags:items at position x_{1,1,1} or x_{1,1} (e.g. gene identifiers)
You have to run filterConCub()
before you can get the table. If filterConCub()
was not run, a warning is shown and getTable
returns NULL
.
signature(object = "concub")
GOStats
to perform a simple two-way enrichment analysis
showClass("concub")