Hi All! I have a table (see attached shortened example). Here the columns represent the samples (two here) and the rows the probe pairs. 16 probe pair represent a probe set (however, some probe sets contain only 11 or 14 probe pairs!). I would like to divide the probe pairs by the average of a probe set. How can I do this? (I am looking for splice variants, therefore the actual expression of the gene is not important, but the expression of a given probe pair compared to the whole gene set. I cannot go back to the raw data as the table represents already pre-processed data.) Thank you: Balazs ---------------------- SAMPLE1 SAMPLE2 1007_s_at1 20 119.1413 1007_s_at2 20 20 1007_s_at3 20 20 1007_s_at4 20 20 1007_s_at5 1122.83448 1503.4413 1007_s_at6 477.13448 1868.9413 1007_s_at7 29.83448 458.9413 1007_s_at8 20 136.6413 1007_s_at9 20 305.3413 1007_s_at10 373.53448 1567.3413 1007_s_at11 340.83448 596.6413 1007_s_at12 68.83448 293.6413 1007_s_at13 184.83448 673.9413 1007_s_at14 20 20 1007_s_at15 20 72.6413 1007_s_at16 20 124.1413 1487_at1 65.33448 20 1487_at2 20 20 1487_at3 20 20 1487_at4 1664.53448 993.9413 1487_at5 1981.13448 1566.8413 1487_at6 20 20 1487_at7 38.53448 20 1487_at8 20 20 1487_at9 20 20 1487_at10 20 20 1487_at11 91.33448 20 1487_at12 581.33448 800.4413 1487_at13 49.03448 200.9413 1487_at14 20 80.6413 -------------------------------------- ------------------------- Balázs GYÖRFFY MD, PhD Children's Hospital Boston Informatics Program Harvard-MIT Health Sciences and Technology 300 Longwood Avenue, Boston, USA Enders 150.6 Tel: +1 617 919 2654 --------------------------------- [[elided Yahoo spam]] [[alternative HTML version deleted]]

Hi Balazs, Does this do what you need? dat <- read.table("clipboard", header=TRUE) dat$pset <- gsub("([0-9])+$", "", rownames(dat)) # probe set means and number of probes dat.sum <- with(dat, aggregate(dat[, 1:2], by=list(pset), FUN="mean")) dat.sum$n <- with(dat, aggregate(dat[, 1], by=list(pset), FUN="length"))[, 2] # scaled intensities dat$SAMPLE1.SCALED <- dat$SAMPLE1/with(dat.sum, rep(SAMPLE1, n)) dat$SAMPLE2.SCALED <- dat$SAMPLE2/with(dat.sum, rep(SAMPLE2, n)) dat -Christos Christos Hatzis, Ph.D. Nuvera Biosciences, Inc. 400 West Cummings Park Suite 5350 Woburn, MA 01801 Tel: 781-938-3830 www.nuverabio.com > -----Original Message----- > From: bioconductor-bounces at stat.math.ethz.ch > [mailto:bioconductor-bounces at stat.math.ethz.ch] On Behalf Of > Dr Balazs Gyorffy > Sent: Monday, May 05, 2008 3:34 PM > To: bioconductor at stat.math.ethz.ch > Subject: [BioC] splicing in Affymetrix > > Hi All! > > I have a table (see attached shortened example). Here the > columns represent the samples (two here) and the rows the > probe pairs. 16 probe pair represent a probe set (however, > some probe sets contain only 11 or 14 probe pairs!). I would > like to divide the probe pairs by the average of a probe set. > > How can I do this? > > (I am looking for splice variants, therefore the actual > expression of the gene is not important, but the expression > of a given probe pair compared to the whole gene set. I > cannot go back to the raw data as the table represents > already pre-processed data.) > > Thank you: > Balazs > > ---------------------- > > SAMPLE1 SAMPLE2 > 1007_s_at1 20 119.1413 > 1007_s_at2 20 20 > 1007_s_at3 20 20 > 1007_s_at4 20 20 > 1007_s_at5 1122.83448 1503.4413 > 1007_s_at6 477.13448 1868.9413 > 1007_s_at7 29.83448 458.9413 > 1007_s_at8 20 136.6413 > 1007_s_at9 20 305.3413 > 1007_s_at10 373.53448 1567.3413 > 1007_s_at11 340.83448 596.6413 > 1007_s_at12 68.83448 293.6413 > 1007_s_at13 184.83448 673.9413 > 1007_s_at14 20 20 > 1007_s_at15 20 72.6413 > 1007_s_at16 20 124.1413 > 1487_at1 65.33448 20 > 1487_at2 20 20 > 1487_at3 20 20 > 1487_at4 1664.53448 993.9413 > 1487_at5 1981.13448 1566.8413 > 1487_at6 20 20 > 1487_at7 38.53448 20 > 1487_at8 20 20 > 1487_at9 20 20 > 1487_at10 20 20 > 1487_at11 91.33448 20 > 1487_at12 581.33448 800.4413 > 1487_at13 49.03448 200.9413 > 1487_at14 20 80.6413 > > -------------------------------------- > > > ------------------------- > Balazs GYVRFFY MD, PhD > Children's Hospital Boston Informatics Program Harvard-MIT > Health Sciences and Technology 300 Longwood Avenue, Boston, > USA Enders 150.6 > Tel: +1 617 919 2654 > > > --------------------------------- > [[elided Yahoo spam]] > [[alternative HTML version deleted]] > >