stepNorm (version 1.44.0)

fitWithin: Simple location normalization function for cDNA microarray data

Description

This function performs location normalization on cDNA micoroarray. It operates on class marrayRaw or class marrayNorm. It allows the user to choose from a set of three basic normalization procedures.

Usage

fitWithin(x.fun = "maA", y.fun = "maM", z.fun = TRUE, subset=TRUE, fun = "medfit", ...)

Arguments

x.fun

Name of accessor method for spot intensity, usually maA.

y.fun

Name of accessor method for spot statistics, usually the log-ratio maM.

z.fun

Name of accessor method for spot statistic used to stratify the data, usually a layout parameter, e.g. maPrintTip or maCompPlate. If z is not a character, e.g. NULL, the data are not stratified.

subset

A "logical" or "numeric" vector indicating the subset of points used to compute the normalization values.

fun

Character string specifying the normalization procedure:

medfit

for global median location normalization

rlmfit

for global intensity or A-dependent location normalization using the rlm function

loessfit

for global intensity or A-dependent location normalization using the loess function

...

Miscs arguments to be passed in fun

Value

The function fitWithin returns a function($F$) with bindings for x.fun, y.fun, z.fun, subset and fun. When the function $F$ is evaluated with an object of class marrayNorm or marrayRaw, it carries out normalization and returns an object of class marrayFit that contains the normalization information as a list with the following list components:

varfun

: A character vector of names of predictor variables.

x

: A numeric matrix of predictor variables.

y

: A numeric matrix of repsonses.

residuals

: A numeric matrix of normalized values (typically log ratios ($M$)).

fitted

: A numeric matrix of the fitted values.

enp

: The equivalent number of parameters; see loess.

df.residual

: The residual degrees of freedom.

fun

: A character string indicating the name of the function used for normalization.

Note that the residuals component stores the normalized ratios.

Details

Normalization is typically performed on the expression ratios of cDNA microarray data, using the function specified by argument fun. Currently, this function is to be chosen from: medfit (median), rlmfit (rlm) and loessfit(loess). When z.fun is provided as a character string, for example, maPrintTip, the normalization procedure is operated within each print-tip of the slide.

References

Y. H. Yang, S. Dudoit, P. Luu, and T. P. Speed (2001). Normalization for cDNA microarray data. In M. L. Bittner, Y. Chen, A. N. Dorsel, and E. R. Dougherty (eds), Microarrays: Optical Technologies and Informatics, Vol. 4266 of Proceedings of SPIE.

See Also

fit2DWithin

Examples

## using the swirl data as example
data(swirl)

## median normalization
med <- fitWithin(fun="medfit")
swirl1.med <- med(swirl[,1])
norm.M <- swirl1.med$residuals ## matrix of normalized ratios

## rlm normalization
rlmF <- fitWithin(fun="rlmfit")
swirl1.rlm <- rlmF(swirl[,1])

## loess normalization, default span=0.4
loessF <- fitWithin(fun="loessfit")
swirl1.loess <- loessF(swirl[,1])
## loess normalization, span=0.2
loessF.1 <- fitWithin(fun="loessfit", span=0.2)
swirl1.loess.1 <- loessF.1(swirl[,1])


## within-printtip loess normalization
loessP <- fitWithin(z.fun="maPrintTip", fun="loessfit")
swirl1.loessP <- loessP(swirl[,1])