metaseqR (version 1.12.2)

stat.nbpseq: Statistical testing with NBPSeq

Description

This function is a wrapper over NBPSeq statistical testing. It accepts a matrix of normalized gene counts or an S4 object specific to each normalization algorithm supported by metaseqR.

Usage

stat.nbpseq(object, sample.list, contrast.list = NULL,
        stat.args = NULL, libsize.list = NULL)

Arguments

object

a matrix or an object specific to each normalization algorithm supported by metaseqR, containing normalized counts. Apart from matrix (also for NOISeq), the object can be a SeqExpressionSet (EDASeq), CountDataSet (DESeq), DGEList (edgeR) or list (NBPSeq).

sample.list

the list containing condition names and the samples under each condition.

contrast.list

a named structured list of contrasts as returned by make.contrast.list or just the vector of contrasts as defined in the main help page of metaseqr.

stat.args

a list of NBPSeq statistical algorithm parameters. See the result of get.defaults("statistics", "nbpseq") for an example and how you can modify it. It is not required when the input object is already a list from NBPSeq normalization as the dispersions are already estimated.

libsize.list

an optional named list where names represent samples (MUST be the same as the samples in sample.list) and members are the library sizes (the sequencing depth) for each sample. If not provided, the default is the column sums of the gene.counts matrix.

Value

• A named list of p-values, whose names are the names of the contrasts.

Examples

require(DESeq)
data.matrix <- counts(makeExampleCountDataSet())
sample.list <- list(A=c("A1","A2"),B=c("B1","B2","B3"))
contrast <- "A_vs_B"
norm.data.matrix <- normalize.nbpseq(data.matrix,sample.list)
p <- stat.nbpseq(norm.data.matrix,sample.list,contrast)