"nice separation with PM significantly higher than MM" - that might make the MMs more attractive as candidate controls to correct for cross hybridization, but the real question is whether the resulting corrected PM intensity is proportional to target transcript abundance? This is not automatic. When comparing different ways to isolate RNA from whole blood, for example, we see that the different protocols lead to clear differences in % present calls (i.e. PM significantly higher than MM), but it is not clear whether these differences translate into better measurements of expression (we are conducting an experiment to assess this). Is there any indication that the new labelling kit will lead to better overlall data, or is the change primarily intended to make the PM-MM measures more attractive? -francois ----- Original Message ----- From: "Matthew Hannah" <hannah@mpimp-golm.mpg.de> To: <bioconductor@stat.math.ethz.ch> Sent: Wednesday, September 08, 2004 9:26 AM Subject: [BioC] Affymetrix is changing its labelling kit - data analysiswill change!? > Hi, > > I thought I'd post this here, something to think about, especially if data from the latest chips lands on your desk. > > Had a seminar today from the affy rep - they are changing labelling kit/procedure from ENZO to IVT. The new one is better BUT you will not be able to directly compare between new and old. You will still be able to buy direct from ENZO to complete current projects. > > The biggest change appears to be they will switch to just using biotin-pseudouridine instead of biotin U and C. There are details below, but I can't find the really interesting figure he showed - which was a comparison of the PM and MM signals (of the spike-ins I guess) with ENZO (big overlap - as we know!) and with IVT (nice separation with PM significantly higher than MM). This will probably have a big impact on computing expression and detection. > > Something to keep us on our toes... > > http://www.affymetrix.com/support/technical/technotes/ivt_technote.pdf > > Also, better order the new computer as there could be alot of data - they are already down to 11?m from the previous 18?m, 5?m is on the way, and there is rumour of 1-2?m... > > Cheers, > Matt > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor >