filter.probes {Agi4x44PreProcess}R Documentation

Filtering out probes

Description

Filter out probes according to their Quality Flag

Usage

filter.probes(ddNORM,control = NULL, wellaboveBG = NULL, isfound = NULL, wellaboveNEG = NULL, sat = NULL, PopnOL = NULL, NonUnifOL = NULL, nas = NULL, limWellAbove, limISF, limNEG, limSAT, limPopnOL, limNonUnifOL, limNAS, makePLOT, annotation.package, flag.counts = NULL, targets)

Arguments

ddNORM An RGList in log2 scale to be FILTERED out according to a Quality FLAG
control LOGICAL: If True it removes controls
wellaboveBG LOGICAL: If True it filters by Well Above BG FLAG
isfound LOGICAL: If True it filters by Is Found FLAG
wellaboveNEG LOGICAL: If True it filters by Well Above NEG CTRLS
sat LOGICAL: If True it filters by Is Saturated FLAG
PopnOL LOGICAL: If True it filters by Population Outlier FLAG
NonUnifOL LOGICAL: If True it filters by Non Uniform Outlier FLAG
nas LOGICAL: If True it removes NAs
limWellAbove for a given spot xi accros samples, is the minimum in a experimental condition with a wellaboveBG-FLAG = 1 (Is Well Above BG)
limISF for a given feature xi accros samples, is the minimum in a experimental condition with a isfound-FLAG = 1 (Is Found)
limNEG for a given feature xi accros samples, is the minimum in a experimental condition with a intensity > Limit established for negative controls (Mean + 1.5 x SD)
limSAT for a given feature xi accros samples, is the minimum in a experimental condition with a saturation-FLAG = 0 (Non Saturated)
limPopnOL for a given feature xi accros samples, is the minimum in a experimental condition with a saturation-FLAG = 1 (Is Pop OL)
limNonUnifOL for a given feature xi accros samples, is the minimum in a experimental condition with a saturation-FLAG = 1 (Is Non Uni OL)
limNAS for a given feature xi accros samples, is the minimum in a experimental condition
makePLOT LOGICAL: If True it makes QC graphs filtering
annotation.package a character specifying the AGI annotation package: 'hgug4112a.db','mgug4122a.db'
flag.counts LOGICAL: If True it runs the countFLAG function
targets data.frame with the target structure

Details

Agilent Feature Extraction software provides a flag for each spot that identifies different quantification errors of the signal. Quantification flags were used to filter out signals that did not reach a minimum established criterion of quality. Data were filtered at a probe level according to the following criteria. a- To keep features within the dynamic range of the scanner: For a spot = xi accross all the samples, we demand that at least p probes of the spot xi in at least one experimental condition had a quantification flag denoting that the signal is distinguishable from background. The same criterion is applied independently for the 'IsFound' flag and for signal saturation. b- To keep good quality features, we filtered out probes that had more than y in at least one experimental condition flagged as Outliers.

Value

The function returns a RGList containing with the FILTERED data eliminated In order to allow the tracking of features that may have been filtered out from the original raw data, the following files are given:
RawDataNOCtrl.txt: contains all the features included in the array once the internal controls have been removed. Internal controls are removed prior to any preprocessing step.
IsNOTWellAboveBG.txt: contains the features that have been filtered out because they are not distinguishable from the local background signal.
We uses a Boolean flag indicating if a feature is WellAbove Background (Flag = 1) or not (Flag = 0). A feature reaches a Flag = 1 if IsPosAndSignif and additionally the gBGSubSignal is greater than 2.6*g(r)BG_SD.
IsPosAndSignif uses a Boolean flag, established via a 2-sided t-test, indicates whether the mean signal of a feature is greater than the corresponding background. 1 indicates feature is positive and significant above background
IsNOTFound.txt: contains features that have been filtered out because were NOT FOUND. A feature is considered Found if two conditions are true: 1- the difference between the feature signal and the local background signal is more than 1.5 times the local background noise and 2- the spot diameter is at least 0.30 times the nominal spot diameter. A Boolean variable is used to flag found features. 1 = IsFound
IsSaturated.txt: contains the features that are saturated. A feature is saturated IF 50 threshold. 1 = Saturated
IsFeatNonUnifOL.txt: contains the features that are considered Non Uniformity Outlier. A feature is non-uniform if the pixel noise of feature exceeds a threshold established for a uniform feature. 1 indicates Feature is a non-uniformity outlier.
IsFeatPopnOL.txt: contains the features that are considered Population Outlier. A feature is a population outlier if its signal intensity is lower than a lower threshold or exceeds an upper threshold determined using a multiplier (1.42) times the interquartile range of the population. 1 indicates Feature is a population outlier
IsNOTWellAboveNEG.txt: Besides, for each feature we can demand a minimum signal value that have to be reached at least for a p in one of the experimental conditions. The minimum limit has been established as Mean Negative Controls + 1.5*(Std. dev.Negative Controls). Normally, after filtering by WellAboveBG and IsFound criteria, all probes are well above negative controls.
In addition to all files indicated above we have added ACCNUM, GENE SYMBOL, ENTREZID reference and gene DESCRIPTION corresponding to each manufacturer probe code in the corresponding annotation package.

Author(s)

Pedro Lopez-Romero

Examples

## Not run: 
         data(dd)
         data(targets)
         library(hgug4112a.db)
         ddNORM=BGandNorm(dd,BGmethod='half',NORMmethod='quantile',
                        foreground='MeanSignal',background='BGMedianSignal',
                        offset=50,makePLOTpre=FALSE,makePLOTpost=FALSE)
         ddFILT=filter.probes(ddNORM,
                control=TRUE,
                wellaboveBG=TRUE,
                isfound=TRUE,
                wellaboveNEG=TRUE,
                sat=TRUE,
                PopnOL=TRUE,
                NonUnifOL=TRUE,
                nas=TRUE,
                limWellAbove=75,
                limISF=75,
                limNEG=75,
                limSAT=75,
                limPopnOL=75,
                limNonUnifOL=75,
                limNAS=100,
                makePLOT=TRUE,annotation.package="hgug4112a.db",flag.counts=TRUE,targets)
## End(Not run)

[Package Agi4x44PreProcess version 1.2.0 Index]