[BioC] outliers in Bead summ data

[Matt Ritchie]

Dear Matt,

The createBeadSummaryData() function in beadarray automatically removes 
outliers and summarises the intensities from the replicate beads when you 
specify method="illumina". This is done using Illumina's 3 MADs from the 
median rule to determine outliers by default. It is worth finding out 
whether outliers have already been removed from your bead-level data to 
avoid doing this step twice. If BeadScan has been run with the line

<ExcludeOutliers>true</ExcludeOutliers>

in the settings.xml file (located in the directory where BeadScan is 
installed), then you should use method="mean" instead, since the outliers 
will have already been removed from the bead-level output. Ask someone in 
your core facility to check on this if you're unsure. If you want the 
outliers included in your output, changing this 'true' to a 'false' should 
work.

I hope this helps.  Best wishes,

Matt

>Hi all,
>
>Does anyone have a function that is able to remove outliers from bead 
>summary data? I am not a bioinformatian or a programmer, and I have no 
>idea on how to change or alter the function used on Bead image data.
>
>I.e:
>par(mfrow=c(2,5))
>for(i in 1:10){
>  o=findAllOutliers(BLData.bc, array=i)
>  plotBeadLocations(BLData.bc, array=i, BeadIDs=o, main=an[i], SAM=TRUE, 
>pch=".")
>}
>
>outliers = NULL
>for(i in 1:10) {
>  outliers[i] = length(findAllOutliers(BLData.bc, array=i))
>}
>x11()
>par(mai=c(2,1,0.2,0.1))
>barplot(outliers/numBeads(BLData.bc)*100, main="Outliers per array", 
>ylab="%", las=2, names=an)
>
>Thanks,
>Matt