BMEA: Bayesian Modelling for Exon Arrays

Description Usage Arguments Details Value Examples

Based on the background priors for each probe, the Z-score is a measure of whether there is any detectable signal above background (DABG)

1	zScore(PM, lambda, delta, exons = NULL)

`PM`	the raw observed intensities for the given unit. Must be a matrix with dimensions nrows=nProbes (K) & ncol= nChips (I).
`lambda`	the vector of means for the log-normal distribution of the background signal.
`delta`	the vector of standard deviations for the log-normal distribution of the background signal.
`exons`	the exon structure of the gene as factors. If supplied, the function will return a value for each exon.

If no signal is present for a group of probes, any observed intensites will be samples from the distribution for background signal. Genes with z-score below a pre-defined threshold (e.g. 1.96) can be declared as not expressed, and likewise for any given exon.

If cacluating the value on a cumulative basis, individual scores will first be calculated for each exon, and then on an increasing basis.

A list with the components:

$gene the Z-Score for the entire gene
$exons the Z-Score for each exon

If the exons vector is not supplied, the $exons component will return NULL.

## Assuming the PM matrix, and the vectors of means & sds
## have already been obtained
PM <- matrix(exp(rnorm(120, 4, 2)), nrow=12, ncol=10) # A matrix of BG only pseudo-PM values
lambda <- matrix(4, nrow=12, ncol=10) # E[BG] at each probe
delta <- matrix(2, nrow=12, ncol=10) # SD[BG] at each probe
exons <- as.factor(rep(paste("Exon",1:3, sep=""), each=4)) # The exon structure
z <- zScore(PM, lambda, delta, exons) # The zScore
z