interactiontable: Returns a list of interactions with associated statistics.

Description

This is a extanded wrapper around the topTable function from the limma package, as an option the ordinary t statistics can be calculated as well.

Usage

interactiontable(ebfit, sort = "none", ord.t = FALSE, correction = "BH")

Arguments

ebfit

ebfit a MArrayLM object produced by the eBayes function

sort

character string specifying which statistic to rank genes by, possible arguments are none, ID,size, t,B,adj.P.val,P.Value, and if ord.t = TRUE: ord.t, ord.p and ord.p.adj.

ord.t

Logical, should ordinary t statistics be calculted? Default is FALSE.

correction

method used to adjust the p-values for multiple testing. Default is BH. See p.adjust for the complete list of options.

Value

Returns a dataframe where the rows are the interaction pairs and the columns the statistics:ID: Interaction pair ifsize: the average interaction sizet: the moderated t statisticsP.Value: p-value for the moderated t statisticsadj.P.Val: adjusted p-valueB: the b statisticsif the ord.t=TRUE, the ordinary t statistics (ord.t), with correspnding p-values (ord.p) and adjusted p-values (ord.p.adj)

Warning

usage of the ordinary t statistics is not recommended for data sets with few replicates.

Examples

Run this code


## simulated data
     y <- matrix(rnorm(50*4,sd=1),50,4)
     rownames(y) <- paste("Pair",1:50)
    
     # fit and eBayes
     fit <- lmFit(y)
     fit <- eBayes(fit)
     tt = interactiontable(fit,sort="size")
     head(tt)