> Hi Mark, > > I was able to run RP from the M values but have couple of doubts. Thank you for suggesting that to me. I really appreciate if you could help me out with it. First question - is it possible to use only significantly differentially expressed genes obtained from the linear model (LIMMA ) for RP analysis. I was unable to select the subset of genes from topTable command. Second question is something very simple but I am unable to get it to work. I want to see the gene ID/gene names in the RP output tables. For some reason it is unable to recognize the gene names column from the file. Session info is pasted below. > > Thank you, Mark. > > > Sincerely, Neel > > > Session info > > > library(limma) > > getwd() > [1] "/Users/Neel" > > setwd("/Users/Neel/agilent/limma") > > getwd() > [1] "/Users/Neel/agilent/limma" > > targets = readTargets() > > targets = readTargets("Targets.txt", row.names = "Name") > > spottypes = readSpotTypes() > > myfun = function(x) as.numeric(x$ControlType > -50.5) > > RG = read.maimages(targets, source="agilent", wt.fun=myfun) > Read conta.txt > Read contb.txt > Read contc.txt > Read contd.txt > Read pcba.txt > Read pcbb.txt > Read pcbc.txt > Read pcbd.txt > > spottypes = readSpotTypes() > > RG$genes$Status = controlStatus(spottypes, RG) > Matching patterns for: ProbeName GeneName > Found 42990 gene > Found 14 BLANK > Found 604 Blank > Found 320 positive > Found 153 negative > Found 130 flag1 > Found 120 flag2 > Setting attributes: values Color > > RG.b = backgroundCorrect(RG, method="normexp", offset=50) > Green channel > Corrected array 1 > Corrected array 2 > Corrected array 3 > Corrected array 4 > Corrected array 5 > Corrected array 6 > Corrected array 7 > Corrected array 8 > Red channel > Corrected array 1 > Corrected array 2 > Corrected array 3 > Corrected array 4 > Corrected array 5 > Corrected array 6 > Corrected array 7 > Corrected array 8 > > MA.p = normalizeWithinArrays(RG.b, method="loess") > > MA.pAq = normalizeBetweenArrays(MA.p, method = "Aquantile") > > design = cbind(CONT=c(0,0,0,0,1,1,1,1), PCB=c(1,1,1,1,0,0,0,0)) > > isGene = RG$genes$Status=="gene" > > fit = lmFit(MA.p[isGene,], design) > > cont.matrix = makeContrasts(PCBvsCONT= CONT-PCB, levels=design) > > fit2 = contrasts.fit(fit, cont.matrix) > > fit2 = eBayes(fit2) > > top1 = topTableF(fit2, number=300, genelist=fit$genes, adjust.method="BH", sort.by="F", p.value=1) > > write.table(top1, file= "Fstatistic.txt", quote=FALSE, sep="\t", row.names=FALSE, col.names=TRUE) > > cl = c(rep(0,4), rep(1,4)) > > RP.out = RP(MA.p, cl, logged=TRUE, rand=123) > Rank Product analysis for two-class case > > Starting 100 permutations... > Computing pfp .. > Outputing the results .. > > topGene (RP.out, cutoff=0.05, logged = TRUE) > Table1: Genes called significant under class1 < class2 > > Table2: Genes called significant under class1 > class2 > > $Table1 > gene.index RP/Rsum FC:(class1/class2) pfp P.value > [1,] 25521 1.1388 0.0296 0.0000 0 > [2,] 26068 1.8476 0.0335 0.0000 0 > [3,] 21312 153.5698 0.3200 0.0000 0 > [4,] 2577 303.9498 0.3469 0.0200 0 > [5,] 6418 330.7369 0.3833 0.0200 0 > [6,] 7471 338.5279 0.3641 0.0167 0 > [7,] 31524 415.9073 0.3912 0.0314 0 > > $Table2 > gene.index RP/Rsum FC:(class1/class2) pfp P.value > [1,] 4970 138.9352 3.5370 0.0000 0 > [2,] 6578 152.4629 3.5384 0.0000 0 > [3,] 9065 177.9296 3.3864 0.0000 0 > [4,] 8198 336.6878 2.9670 0.0200 0 > [5,] 5300 369.8948 2.8839 0.0280 0 > [6,] 15935 393.9698 2.7581 0.0350 0 > [7,] 7262 445.7020 2.5756 0.0486 0 > [8,] 3089 471.5363 2.6675 0.0488 0 > > names(MA.p) > [1] "weights" "targets" "genes" "source" "M" "A" > > topGene (RP.out, num.gene = 10, logged = TRUE, gene.names = MA.p$genes) > Warning: gene.names should have the same length as the gene vector. > No gene.names are assigned > > > On Dec 17, 2009, at 7:09 PM, Mark Cowley wrote: > >> no worries Neel, >> you should be able to run RP using the M values, ie the R-G log2 ratios that you fitted the linear models to. >> mark >> >> On 17/12/2009, at 2:07 PM, Neel Aluru wrote: >> >>> Hi Mark, >>> >>> Thanks for your mail. I was able to figure out how to get the replicate values. However, I haven't tried working on rank product analysis yet. Will get to it in a couple of days. Thanks a lot for enquiring about my progress. The mailing list is really helpful to learn R packages. >>> >>> Sincerely, Neel >>> >>> At 07:57 PM 12/16/2009, you wrote: >>>> hi Neel, >>>> how did you go with this problem? let me know if you still need it >>>> answering >>>> mark >>>> On 11/12/2009, at 1:45 PM, Neel Aluru wrote: >>>> >>>>> Hello, >>>>> >>>>> I am hoping someone will be able to help me with this. I am >>>>> analyzing the two color common reference design agilent arrays using >>>>> Limma. I have four replicates and two treatments(control and PCB). I >>>>> am trying to extract fold change values (PCB/control) for the four >>>>> replicates separately. lmfit() command after normalizing between >>>>> arrays is pooling all the samples and I get one fold change value. >>>>> Is there anyway I can get the values for individual replicates >>>>> separtely? I want to use the replicate values in Rank Product >>>>> analysis. Any help will be really helpful. >>>>> >>>>> Sorry for bothering everyone with my questions. I have made a lot of >>>>> progress since I started using this package and it is mainly due to >>>>> this mailing list. >>>>> Sincerely, Neel >>>>> >>>>> >>>>> Neel Aluru >>>>> Postdoctoral Scholar >>>>> Biology Department >>>>> Woods Hole Oceanographic Institution >>>>> Woods Hole, MA 02543 >>>>> USA >>>>> 508-289-3607 >>>>> >>>>> _______________________________________________ >>>>> Bioconductor mailing list >>>>> Bioconductor@stat.math.ethz.ch >>>>> https://stat.ethz.ch/mailman/listinfo/bioconductor >>>>> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor >>>> >>> Neelakanteswar Aluru Ph.D. >>> Post doctoral Scholar >>> Biology Department >>> Redfield 304 (MS#32) >>> Woods Hole Oceanographic Institution >>> Woods Hole MA 02543 USA >>> Phone: (508) 289-3607 [Office] >>> 774-392-3727 [Cell] >>> RID: A-7237-2009 >>> >> > > Neel Aluru > Postdoctoral Scholar > Biology Department > Woods Hole Oceanographic Institution > Woods Hole, MA 02543 > USA > 508-289-3607 > > > Neel Aluru Postdoctoral Scholar Biology Department Woods Hole Oceanographic Institution Woods Hole, MA 02543 USA 508-289-3607 [[alternative HTML version deleted]]