splicegear (version 1.44.0)

getRelSignStrength: functions to perform SPLICE

Description

Implementations of the SPLICE algorithm

Usage

getRelSignStrength(x, tissue = as.factor(1:ncol(x)), fun = mean, nipt = 30, nitt = 30, ...)
getFinalRatio(x, tissue=as.factor(1:ncol(x)), fun=mean, ...)

Arguments

x

a matrix. One probe per line, one column per sample. Typically this would be the slot exprs of an instance of class ExprSet.

tissue

a covariate (factor) about the samples.

fun

a function to obtain a summary value (mean by default)

nipt

see reference.

nitt

see reference.

...

optional parameters for the function fun

Value

A matrix of the same dimension than the input x, holding 'RSS' (Relative Signal Strength) or 'final ratios' respectively, as described in the reference. Two attributes nip and nit are attached the returned matrix.

Details

getFinalRatio will call getRelSignStrength. The values are log-transformed. It is probably a good idea to avoid feeding function with values that are already on log scale.

References

Genome Research (2001), Hu et. al., vol. 11, p.1244

Examples

data(spliceset)

## The intensity values in the example are log-transformed.
## Undo by taking the exponential
exprs(spliceset) <- exp(exprs(spliceset))

## Re-order the rows of different slots to have the probes sorted by
## position
spliceset <- sort.SpliceExprSet(spliceset)
## extract the expression matrix
expr.m <- exprs(spliceset)
fr <- getFinalRatio(expr.m, tissue=pData(spliceset@eset)[[1]])