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

Description Usage Arguments Details Fields and Methods Note Author(s) Examples

Package: aroma
Class MAData

Object
~~|
~~+--MicroarrayData
~~~~~~~|
~~~~~~~+--MAData

Directly known subclasses:
BMAData, TMAData

public static class MAData
extends MicroarrayData

Creates a new MAData object. The philosophy behind this data structure is to think about the data in the form of log ratios (M) and log intensites (A) for the spot signals. This is in contrast to the idea of the RGData structure, which thinks about the data as the signal in one channel (R) versus the signal in the other channel (G).

1	MAData(M=NULL, A=NULL, layout=NULL, extras=list())

`M,A`	A NxM `matrix` containing (base-2) log-ratios and log-intensities, respectively.
`layout`	A `Layout` object specifying the spot layout of the slides in this data set.
`extras`	Private argument. Do not use.

The mapping between M and A, and R and G is a one-to-one function, i.e. you can go back and forth without loosing any information. Given the signal R and G for the R and the G channels you get the M and the A values by:

M = log2(R/G), A = log2(sqrt(R*G)) = 1/2*log2(R*G),

and going back to the R and the G by:

R = sqrt(2^(2A+M)), G = sqrt(2^(2A-M))

Comments: M is a memnonic for Minus and A is for Add since M = \log_2{R}-\log_2{G} and A = 1/2*(\log_2{R}+\log_2{G}).

Fields

	`M`	The log of the ratios between R and G, i.e. log2(R/G) where R and G is the signal for the R and the G channel.
	`A`	The log of the intensities of R and G, i.e. 1/2log2(RG) where R and G is the signal for the R and the G channel.

Methods:

	`as.character`	-
	`as`	-
	`as.MAData`	-
	`as.RawData`	-
	`as.RGData`	Transform MA format into RG format.
	`boxplot`	-
	`drawCurveFit`	-
	`getColors`	Generates colors for each of the specified spots.
	`getDenseSpots`	-
	`getGeneVariability`	Gets the genewise variability.
	`getHistogram`	-
	`getMOR`	Gets the Measure of Reproducibility.
	`getSpotVariability`	-
	`hist`	-
	`mean`	Average Mean for microarray data.
	`normalizeAcrossSlides`	Normalizes across slides.
	`normalizeAffine`	Affine normalization based on non-logged data.
	`normalizeCurveFit`	Within-slide normalization that adjust log-ratios by estimating a smooth intensity-dependent curve in (A,M).
	`normalizeGenewise`	-
	`normalizePlatewise`	-
	`normalizePrintorder`	-
	`normalizeQuantile`	-
	`normalizeSpatial`	-
	`normalizeWithinSlide`	Within-slide normalization.
	`plot`	-
	`plot3d`	-
	`plotDensity`	-
	`plotDiporder`	-
	`plotMvsM`	Plots the log-ratios for one slide versus another.
	`plotPrintorder`	-
	`plotSpatial`	-
	`plotSpatial3d`	-
	`plotXY`	-
	`qqnorm`	-
	`range`	-
	`shift`	Shift the log-ratios, log-intensities or the raw signal.
	`shiftEqualRG`	-
	`swapDyes`	Dye swap one or many slides.
	`topSpots`	Gets the top spots.
	`var`	Variance for microarray data.

Methods inherited from MicroarrayData:
addFlag, append, applyGenewise, applyGroupwise, applyPlatewise, applyPrintdipwise, applyPrinttipwise, as.character, as.data.frame, boxplot, clearCache, clearFlag, createColors, dataFrameToList, equals, extract, getBlank, getCache, getChannelNames, getColors, getExcludedSpots, getExtra, getExtreme, getFieldNames, getFlag, getInclude, getLabel, getLayout, getProbeWeights, getSignalWeights, getSlideNames, getSlidePairs, getSpotPosition, getSpotValue, getTreatments, getView, getWeights, getWeightsAsString, hasExcludedSpots, hasLayout, hasProbeWeights, hasSignalWeights, hasWeights, highlight, hist, isFieldColorable, keepSlides, keepSpots, listFlags, lowessCurve, nbrOfDataPoints, nbrOfFields, nbrOfSlides, nbrOfSpots, nbrOfTreatments, normalizePlatewise, normalizePrintorder, normalizeQuantile, plot, plotDensity, plotGene, plotPrintorder, plotReplicates, plotSpatial, plotSpatial3d, plotXY, points, putGene, putSlide, qqnorm, quantile, range, range2, read, readHeader, readToList, removeSlides, removeSpots, resetProbeWeights, resetSignalWeights, select, seq, setCache, setExcludedSpots, setExtra, setFlag, setLabel, setLayout, setProbeWeights, setSignalWeights, setSlideNames, setTreatments, setView, setWeights, size, str, subplots, summary, text, updateHeader, validateArgumentChannel, validateArgumentChannels, validateArgumentGroupBy, validateArgumentSlide, validateArgumentSlides, validateArgumentSpotIndex, validateArgumentWeights, write, writeHeader

Methods inherited from Object:
$, $<-, [[, [[<-, as.character, attach, attachLocally, clearCache, clone, detach, equals, extend, finalize, gc, getEnvironment, getFields, getInstanciationTime, getStaticInstance, hasField, hashCode, ll, load, objectSize, print, save

There are several functions that returns an object of this class, and it is only in very special cases that you actually have to create one yourself.

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

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

  # Create a raw data object from the preexisting example data in
  # the sma package.
  SMA$loadData("mouse.data")
  layout <- Layout$read("MouseArray.Layout.dat", path=system.file("data-ex", package="aroma"))
  raw <- RawData(mouse.data, layout=layout)

  # Get the signal (here by default non-background corrected)
  ma <- getSignal(raw, bgSubtract=TRUE)

  # Transform (M,A) into (R,G)
  rg <- as.RGData(ma)

  # Transform back from (R,G) to (M,A)
  ma2 <- as.MAData(rg);

  # Check that the tranformation a one-to-one function
  print(equals(ma, ma2))   # TRUE

  layout(matrix(1:4, ncol=2, byrow=TRUE))
  # Plot the R vs G with a fitted (lowess) line for slide 2.
  plot(rg, slide=2); lowessCurve(rg)

  # And the similar for M vs A.
  plot(ma, slide=2); lowessCurve(ma)

  # Plot a spatial representation of the M's.
  plotSpatial(ma, slide=2)

  # Make a boxplot of the print-tip groups.
  boxplot(ma, groupBy="printtip", slide=2)