[BioC] What should happen to control probe information in bead level Illumina analyses?

[Simon Lin]

Hi Alice,

Please check the expression level of these negative control probes. They 
must be very low. If all the negative control probes are high, you might get 
a hybridizatin problem; check the stringency control probes.

If these low-expressed negative controls are in your Top50 lists, you might 
try a VST transformation method as reported in
http://nar.oxfordjournals.org/cgi/content/abstract/gkm1075v1

Good luck!

Simon Lin
Northwestern
===========================
Date: Fri, 7 Mar 2008 16:48:02 +1300
From: "Johnstone, Alice" <Alice.Johnstone at esr.cri.nz>
Subject: [BioC] What should happen to control probe information in
bead level Illumina analyses?
To: "bioconductor" <bioconductor at stat.math.ethz.ch>
Message-ID:
<92B672A208AB2F419334EE1EC8C58D9D01D13400 at kscmail1.esr.cri.nz>
Content-Type: text/plain; charset="us-ascii"

Hi again!

When running through my analysis of bead summary data that I have
created from bead level data (background corrected with normexp, and
summarised standard) I produced a topTable list which contains probeIDs
that are not found in Illumina's reference file.  When I checked the
controlprobeprofile.txt there they were!
It looks like I haven't got huge DE change in my experiment, but I still
wouldn't have expected Illumina's negative controls to make the top 50..
Somethings not right.