Hi, I have data obtained with the following procedure under R2.0dev / XPPro / PC32bits: triplicates of 4 different conditions under Affy MmU74Av2. cells obtained from mouse. data <- ReadAffy() data2 <- gcrma(data) I used the development version of GCRMA1.1.0 and AFFY downloaded/updated today. and If I look at the distribution of the intensities after gcrma I obtain an usual plot (see attached). I have attached the degradation plot that for me seems to be fine but the ends (any comments on these ?), and the histogram obtained after hist(data) which identify 3 chips with different profiles. I have never worked with gcrma and mouse chips but I m very surprised of the 2-peak plot after GCRMA; is it a problem of saturation ? or can we suspect that all these 12 samples that look all the same had 2 cell populations inside ??? I doubt about that... any other option ? a problem in GCRMA ? second question, shall I avoid using the 3 chips with very different profiles after hist(data) ? they seem to be fine after GCRMA. thanks for your help Philippe

Hi, Philippe, Please check my reply to you earlier today Jean On Thu, 19 Aug 2004 Phguardiol@aol.com wrote: > Hi, > I have data obtained with the following procedure under R2.0dev / XPPro / PC32bits: triplicates of 4 different conditions under Affy MmU74Av2. cells obtained from mouse. > > data <- ReadAffy() > data2 <- gcrma(data) > > I used the development version of GCRMA1.1.0 and AFFY downloaded/updated today. > > and If I look at the distribution of the intensities after gcrma I obtain an usual plot (see attached). I have attached the degradation plot that for me seems to be fine but the ends (any comments on these ?), and the histogram obtained after hist(data) which identify 3 chips with different profiles. > I have never worked with gcrma and mouse chips but I m very surprised of the 2-peak plot after GCRMA; is it a problem of saturation ? or can we suspect that all these 12 samples that look all the same had 2 cell populations inside ??? I doubt about that... any other option ? a problem in GCRMA ? second question, shall I avoid using the 3 chips with very different profiles after hist(data) ? they seem to be fine after GCRMA. > thanks for your help > Philippe >