RmaBackgroundCorrection: The RmaBackgroundCorrection class
Description
Package: aroma.affymetrix
Class RmaBackgroundCorrection
Object
~~|
~~+--AromaTransform
~~~~~~~|
~~~~~~~+--Transform
~~~~~~~~~~~~|
~~~~~~~~~~~~+--ProbeLevelTransform
~~~~~~~~~~~~~~~~~|
~~~~~~~~~~~~~~~~~+--BackgroundCorrection
~~~~~~~~~~~~~~~~~~~~~~|
~~~~~~~~~~~~~~~~~~~~~~+--RmaBackgroundCorrection
Directly known subclasses:
public static class RmaBackgroundCorrection
extends BackgroundCorrection
This class represents the RMA background adjustment function.Usage
RmaBackgroundCorrection(..., addJitter=FALSE, jitterSd=0.2)
Arguments
addJitter
If TRUE, Zero-mean gaussian noise is added to the
signals before being background corrected. jitterSd
Standard deviation of the jitter noise added.
Fields and Methods
Methods:
rll{
process Performs background correction.
}
Methods inherited from BackgroundCorrection:
process
Methods inherited from Transform:
getOutputDataSet, getOutputDataSetOLD20090509, getOutputFiles
Methods inherited from AromaTransform:
getExpectedOutputFiles, getExpectedOutputFullnames, getFullName, getInputDataSet, getName, getOutputDataSet, getOutputDataSet0, getOutputFiles, getPath, getTags, isDone, process, setTags
Methods inherited from Object:
asThis, getChecksum, $, $<-, [[, [[<-, as.character, attach, attachLocally, clearCache, clearLookupCache, clone, detach, equals, extend, finalize, gc, getEnvironment, getFieldModifier, getFieldModifiers, getFields, getInstantiationTime, getStaticInstance, hasField, hashCode, ll, load, objectSize, print, registerFinalizer, saveJitter noise
The fitting algorithm of the RMA background correction model may not
converge if there too many small and discrete signals. To overcome
this problem, a small amount of noise may be added to the signals
before fitting the model. This is an ad hoc solution that seems to
work.
However, add Gaussian noise may generate non-positive signals.Details
Internally bg.adjust is used to background correct the
probe signals. The default is to background correct PM signals only.