normalizePlatewise.MicroarrayData: Normalization performed plate by plate

In HenrikBengtsson/aroma: An R Object-oriented Microarray Analysis package

Description

Performs a normalization plate by plate on each of the slide seperately. For details on plate-wise normalization see [1] and for details on intensity dependent normalization see [2].

Note that the data in the object is replaced with the new normalized data and the old data is removed. To keep the old data, make a copy of the object before normalizing by using clone(ma), see clone.Object and example below.

Usage

## S3 method for class 'MicroarrayData'
normalizePlatewise(this, field, method="constant", slides=NULL, ...)

Arguments

field	The data field to be normalized.
method	The normalization method to be used. If "constant" the values of the data field will be shifted to have zero bias within each plate group. If "A", the values within each plate group will be normalized against intensity dependent effects.
slides	Slides to be included in the normalization. If NULL, all slides are normalized.
...	Other arguments accepted by underlying normalization methods.

Details

The data will be normalize within each plate group individually by either a) assuming same shift and scale for whole plate i) shifting the data to have median/mean zero. ii) rescale the data so all the groups have the same median absolute deviation (MAD) or standard deviation (sd) or by b) assuming intensity dependent effects. For more details see [1].

Author(s)

Henrik Bengtsson (http://www.braju.com/R/)

References

\item

[1]Henrik Bengtsson, Plate Effects in cDNA microarray data, Matemathical Statistics, Centre for Matematical Sciences, Lund University, Sweden. Manuscript, 2002. \item[2]S. Dudoit, Y. H. Yang, M. J. Callow, and T. P. Speed. Statistical methods for identifying differentially expressed genes in replicated cDNA microarray experiments (Statistics, UC Berkeley, Tech Report 578). URL: http://www.stat.berkeley.edu/users/terry/zarray/Html/papersindex.html

See Also

For other within-slide normalization see MAData.normalizeWithinSlide. For across-slide normalization see MAData.normalizeAcrossSlides. For more information see MicroarrayData.

Examples

   # The option 'dataset' is used to annotate plots.
   options(dataset="sma:MouseArray")

   SMA$loadData("mouse.data")
   layout <- Layout$read("MouseArray.Layout.dat", path=system.file("data-ex", package="aroma"))
   raw <- RawData(mouse.data, layout=layout)
   ma <- getSignal(raw)

   # Scaled intensity normalization print-tip by print-tip
   ma.norm1 <- clone(ma)
   normalizeWithinSlide(ma.norm1, method="s")

   # Intensity normalization plate by plate
   ma.norm2 <- clone(ma.norm1)
   normalizePlatewise(ma.norm2, field="M", method="A")

   # Plot data before and after normalization.
   layout(matrix(1:9, ncol=3, byrow=TRUE))
   plot(ma)
   plotSpatial(ma)
   plotPrintorder(ma)
   plot(ma.norm1)
   plotSpatial(ma.norm1)
   plotPrintorder(ma.norm1)
   plot(ma.norm2)
   plotSpatial(ma.norm2)
   plotPrintorder(ma.norm2)
 

HenrikBengtsson/aroma documentation built on May 7, 2019, 12:56 a.m.