MAplotPara: Parallelized relative M vs. A plots for Microarray Data

In affyPara: Parallelized preprocessing methods for Affymetrix Oligonucleotide Arrays

Description

Parallelized creation of M vs A plots. Where M is determined relative to a specified chip or to a pseudo-median reference chip. And optimized plots for a large number of microarrays.

Usage

MAplotPara(object, 
                       log=TRUE, 
                       type=c("both","pm","mm"),
                       ref=NULL,
                       which=NULL, 
                       ref.title="vs mean ref.Array",
                       subset=NULL,
                       span=1/4,
                       show.statistics=TRUE,
                       family.loess ="gaussian", 
                       pchs=".",
                       plot =TRUE,
                       cutoff =0.5,# add parameter to generic function ma.plot
                       level=1,
                       cluster, verbose = getOption("verbose"),
                       ... )
                       

Arguments

object	An object of class AffyBatch OR a character vector with the names of CEL files OR a (partitioned) list of character vectors with CEL file names.
...	Additional parameters for the routine.
log	A logical value. TRUE as default. Computes logarithms of Array's intensities.
type	Defines how the Affimetrix Array's intensities should be considered (only for Affymetrix: perfect match(pm), mismatch or both).
ref	gives the possibility to define some array of the Affybatch object as 'medianchip' otherwise it will be calculate as reference median array.
which	A list of index samples from the object class, which indicates the samples to be ploted.
ref.title	character- gives a 'title' to label the plots by ma.plot function.
subset	A set of indices to use when drawing the loess curve.
span	span to be used for loess fit.
family.loess	"guassian" or "symmetric" as in loess.
show.statistics	A logical value. TRUE as default. If true some summary statistics of the M values are drawn.
pchs	Graphical plotting 'character'. Default Value "." equivalently to pchs = 46 from the function points
plot	A logical value. TRUE as default, MAplots will be drawn otherwise only the "bad" quality arrays will be calculated.
cutoff	numerical between [0.0-1.0]. As default 0.5. It is considered as limit for the sigma parameter, which is one of the three classificators for the 'bad' quality arrays by MAplotsPara.
level	level- numerical - indicates which level of "bad" quality arrays should be plot if plotDraw =TRUE: 1 - only first level "bad" quality will be considered. First level "bad" array quality are the arrays considered as "bad" after the three possible parameter: S, loess, and sigma 2 - first level "bad" quality and second level will be considered. Second level "bad" quality Arrays are the arrays which has been classified as bad after two of the three possible parameter 3 - all levels will be plot : first, second and third. Third level "bad" quality Arrays are the arrays which are considered as "bad" after one of the three parameter.
cluster	A cluster object obtained from the function makeCluster in the SNOW package. For default .affyParaInternalEnv$cl will be used.
verbose	A logical value. If TRUE it writes out some messages. default: getOption("verbose")

Details

MAplotPara is a function based on the generic function ma.plot from affy package. Only the following parameters are original for MAplotPAra: cluster, object, cuttof, plot, level The parameter ref.fn=c("median","mean") is not allowed because it is not possible to calculate the reference median array as parallelized.

MAplotPara is the parallelized function for MA plots of probe intensities. It is a function to check and control the Data quality of the samples using the MA plot method. For serial function an more details see boxplot. This function is optimized for huge numbers of microarray data.

For using this function a computer cluster using the SNOW package has to be started. Starting the cluster with the command makeCluster generates an cluster object in the affyPara environment (.affyParaInternalEnv) and no cluster object in the global environment. The cluster object in the affyPara environment will be used as default cluster object, therefore no more cluster object handling is required. The makeXXXcluster functions from the package SNOW can be used to create an cluster object in the global environment and to use it for the preprocessing functions.

Value

MAplotPara return a list with four elements: values_MAP, loess_y, quality_MAP, results_MAP.

values_MAP	is a matrix which contains the calculated values for all arrays of the object (sampleNames, S, osc_Loess, sigma, var_sigma).
loess_y	contains the y values (50 points) of the loess curve, which are used to calculate the osc_Loess from values_MAP
quality_MAP	list which contains all "bad" quality Arrays
results_MAP	summary from quality_MAP as matrix, which contains only the Arrays that are considered as "bad" quality and in which levels are they classified. Possible values are 0 if the Array is not at this levels and 1 if it is classified as "bad" sample at this level.

Author(s)

Esmeralda Vicedo <[e-mail]>, Markus Schmidberger [e-mail], Ulrich Mansmann [e-mail]

Examples

## Not run: 
library(affyPara)
if (require(affydata)) {
  data(Dilution)

  makeCluster(3)

  ##MA of Dilution data (affybatch)
  ma1 <- MAplotPara(Dilution)

  ## MAplot to a pdf file
  pdf(file="maplot.pdf", title="AffyBatch MAplot")
  ma2 <- MAplotPara(Dilution)
  dev.off()

  stopCluster()
}

## End(Not run)

affyPara documentation built on Nov. 8, 2020, 11:08 p.m.