Dear Jim, Thanks for pointing out the silly mistake. Sorry for the trouble. Thanks. Warm regards, k On 29/09/2010 21:14, James W. MacDonald wrote: > Hi Karthi, > > On 9/29/2010 3:19 AM, Karthi Sivaraman wrote: >> Dear List, >> >> I am sure this has been discussed before in the list, so I would be >> happy even if someone points me to the right previous discussion. >> >> I have a Nimblegen gene expression dataset as .pairs file. Using dummy >> 2nd channel reading, I have managed to normalize this dataset (vsn >> normalization). lmFit also works with the normalized dataset. However, >> when I try to perform eBayes fit, I am hitting a snag. eBayes fit >> returns a lot of error at this last stage. >> >> I have given the steps / results from each step in this workflow. If >> someone has already faced this issue and resolved it - i would be happy >> to try it. Also, if there is a basic mistake I am making, can someone >> please let me know. Thanks. -- k. >> ===========================================================* >> > library(limma) >> > target=readTargets('list') >> > >> x=read.maimages(target$FileName,columns=list(R="PM",G="PM"),annotat ion=c("PROBE_ID","X","Y")) >> >> Read GSM519643.pair.val >> Read GSM519644.pair.val >> Read GSM519645.pair.val >> Read GSM519646.pair.val >> Read GSM519647.pair.val >> Read GSM519648.pair.val >> Read GSM519649.pair.val >> Read GSM519650.pair.val >> Read GSM519651.pair.val >> > x$R=NULL >> > xNorm=backgroundCorrect(x,method="normexp") >> > library(arrayQualityMetrics) >> Loading required package: affyPLM >> Loading required package: affy >> Loading required package: Biobase >> >> Welcome to Bioconductor >> >> Vignettes contain introductory material. To view, type >> 'openVignette()'. To cite Bioconductor, see >> 'citation("Biobase")' and for packages 'citation(pkgname)'. >> >> Loading required package: gcrma >> Loading required package: preprocessCore >> >> Attaching package: 'affyPLM' >> >> >> The following object(s) are masked from package:stats : >> >> resid, >> residuals, >> weights >> >> > xEsetNorm=new("ExpressionSet",exprs=xNorm$G) >> > library(vsn) >> > xEsetVsn=vsn(xEsetNorm) >> Note: >> The function 'vsn' has been superseded by 'vsn2'. >> The function 'vsn' remains in the package for backward compatibility, >> but for new projects, please use 'vsn2'. > > You might read (and follow) the note... > >> >> vsn: 389307 x 9 matrix (1 stratum). 100% done. >> > design=model.matrix(~ -1+factor(c(1,1,1,2,2,2,3,3,3,)) >> + ) >> Error in factor(c(1, 1, 1, 2, 2, 2, 3, 3, 3, )) : >> element 9 is empty; >> the part of the args list of 'c' being evaluated was: >> (1, 1, 2, 2, 2, 3, 3, 3, ) >> > design=model.matrix(~ -1+factor(c(1,1,1,2,2,2,3,3,3))) >> > colnames(design)=c('Ctrl','Chlr1','Chlr2') >> > design >> Ctrl Chlr1 Chlr2 >> 1 1 0 0 >> 2 1 0 0 >> 3 1 0 0 >> 4 0 1 0 >> 5 0 1 0 >> 6 0 1 0 >> 7 0 0 1 >> 8 0 0 1 >> 9 0 0 1 >> attr(,"assign") >> [1] 1 1 1 >> attr(,"contrasts") >> attr(,"contrasts")$`factor(c(1, 1, 1, 2, 2, 2, 3, 3, 3))` >> [1] "contr.treatment" >> >> > xFit=lmFit(xEsetVsn,design) >> > contmat=makeContrasts(Chlr1-Ctrl,Chlr2-Ctrl,levels=design) >> > contmat >> Contrasts >> Levels Chlr1 - Ctrl Chlr2 - Ctrl >> Ctrl -1 -1 >> Chlr1 1 0 >> Chlr2 0 1 >> > xFit2=eBayes(xFit >> xFit >> > xFit2=eBayes(xFit,contmat) > > There are no arguments for eBayes() that take a contrasts matrix, so I > am not sure what you are trying to do here. You want to do the usual: > > xFit2 <- contrasts.fit(xFit, contmat) > xFit2 <- eBayes(xFit2) > > Best, > > Jim > > > >> Error in log(proportion/(1 - proportion)) - log(r)/2 : >> non-conformable arrays >> In addition: Warning messages: >> 1: In if (ntarget< 1) return(NA) : >> the condition has length> 1 and only the first element will be used >> 2: In if (ntarget< 1) return(NA) : >> the condition has length> 1 and only the first element will be used >> 3: In if (ntarget< 1) return(NA) : >> the condition has length> 1 and only the first element will be used >> 4: In ebayes(fit = fit, proportion = proportion, stdev.coef.lim = >> stdev.coef.lim) : >> Estimation of var.prior failed - set to default value >> 5: In log(proportion/(1 - proportion)) : NaNs produced >> *================================================================== ======= >> >> > -- ======================================= From Ignorance to Truth, From Darkness to Light, From Mortality to Immortality ...

Hi, I wonder if our core facility even knows about that option! Thanks for bringing this up cheers, mark On 30/09/2010, at 11:57 PM, Segal, Corrinne wrote: > Hi, > > If the data is extracted with the FE report set to 'Full' rather > than 'Compact', then it reports the gMeanSignal and gBGUsed (but not > chr_coord). You can then follow the package using the amendments > Pedro posted to get around not having the chr_coord column. > > Cheers, > > Corrinne > > -----Original Message----- > From: bioconductor-bounces at stat.math.ethz.ch [mailto :bioconductor-bounces at stat.math.ethz.ch > ] On Behalf Of David Ruau > Sent: 29 September 2010 19:02 > To: Mark Cowley > Cc: bioconductor at stat.math.ethz.ch > Subject: Re: [BioC] AgiMicroRna problems > > Hi Mark, > > I just received a data set of Human miRNA V3 and AgiMicroRna does > not work. > Basically the column gMeanSignal, gBGUsed, chr_coord are not present. > I modified readMicroRnaAFE accordingly to the post of Pedro (see > after the text) > > My question is why those columns are not present in the txt file. > In the folder I received from our facility there is a XML file > containing the settings of the FeatureExtraction software. One flag > is TextOutPkgType="Compact" > Is there a way to test if this option can be change and what is the > effect on the txt output? > > readMicroRnaAFE <- function (targets, verbose = FALSE) > { > if (!is(targets, "data.frame")) { > stop("'targets' must be a data.frame") > } > ddaux=read.maimages(files=targets$FileName,source="agilent", > other.columns=list(IsGeneDetected="gIsGeneDetected", > IsSaturated ="gIsSaturated", IsFeatNonUnifOF ="gIsFeatNonUnifOL", > IsFeatPopnOL ="gIsFeatPopnOL", BGKmd ="gBGMedianSignal"), > columns=list(R="gTotalGeneSignal", G="gTotalProbeSignal", > Rb="gTotalGeneSignal", Gb="gProcessedSignal"), > verbose=TRUE,sep="\t",quote="" > ) > #return(ddaux) > dd = new("RGList") > dd$R = ddaux$R > dd$G = ddaux$G > dd$Rb = ddaux$Rb > dd$Gb = ddaux$Gb > dd$targets = ddaux$targets > ## suppress column 6 that should have contain chr_pos I guess > dd$genes = ddaux$genes[, c(4, 5)] > dd$other = ddaux$other > rm(ddaux) > if (verbose) { > cat("", "\n") > cat(" RGList:", "\n") > cat("\tdd$R:\t\t'gTotalGeneSignal' ", "\n") > cat("\tdd$G:\t\t'gTotalProbeSignal' ", "\n") > cat("\tdd$Rb:\t\t'gMeanSignal' ", "\n") > cat("\tdd$Gb:\t\t'gProcessedSignal' ", "\n") > cat("", "\n") > } > return(dd) > } > > > David > > On Sep 28, 2010, at 11:52 PM, Mark Cowley wrote: > >> Has anyone had success using AgiMicroRna recently? what array types >> were you using? >> cheers, >> Mark >> >> On 21/09/2010, at 9:44 PM, Mark Cowley wrote: >> >>> Dear Pedro, and BioCers >>> similar to these 2 posts, i'm having problems running AgiMicroRna, >>> because my Agilent TXT files are missing these three columns: >>> gMeanSignal, gBGUsed, chr_coord. >>> https://www.stat.math.ethz.ch/pipermail/bioconductor/2010-August/0 35136.html >>> http://comments.gmane.org/gmane.science.biology.informatics.conduc tor/28101 >>> >>> Here was my first attempt >>>> library("AgiMicroRna") >>>> targets.micro=readTargets(infile="/Volumes/****/projects/****/ >>> targets.txt") (sorry - paranoid collaborator) >>>> dd.micro=readMicroRnaAFE(targets.micro,verbose=TRUE) >>> Error in readGenericHeader(fullname, columns = columns, sep = sep) : >>> Specified column headings not found in file >>> >>> I then tried to recreate my own readMicroRnaAFE which constructed >>> dummy chr_coord, BGKus objects, but then I wasn't able to run the >>> cvArray function: >>>> library("AgiMicroRna") >>>> targets.micro=readTargets(infile="/Volumes/external/projects/LW/ >>> targets.txt") >>>> dd.micro=readMicroRnaAFE(targets.micro,verbose=TRUE) >>> # QC plots ran OK >>>> cvArray(dd.micro,"MeanSignal",targets.micro,verbose=TRUE) >>> Foreground: MeanSignal >>> >>> FILTERING BY ControlType FLAG >>> >>> RAW DATA: 15739 >>> Error in object$other[[k]][i, , drop = FALSE] : >>> incorrect number of dimensions >>>> cvArray(dd.micro,"ProcessedSignal",targets.micro,verbose=TRUE) >>> Foreground: ProcessedSignal >>> >>> FILTERING BY ControlType FLAG >>> >>> RAW DATA: 15739 >>> Error in object$other[[k]][i, , drop = FALSE] : >>> incorrect number of dimensions >>> >>> I gave up on this approach, and instead I followed Pedro's advice in >>> the first URL that I mentioned, and used gTotalSignal instead of >>> gMeanSignal, and removed instances of chr_coord and gBGUsed, but >>> then >>> I can't get TGS, or RMA normalization to work >>> >>>> library("AgiMicroRna") >>>> targets.micro=readTargets(infile="/Volumes/****/projects/****/ >>> targets.txt") (sorry - paranoid collaborator) >>> ddaux=read.maimages(files=targets.micro$FileName,source="agilent", >>> + >>> other.columns=list(IsGeneDetected="gIsGeneDetected", >>> + >>> IsSaturated >>> ="gIsSaturated", >>> + >>> IsFeatNonUnifOF >>> ="gIsFeatNonUnifOL", >>> + >>> IsFeatPopnOL >>> ="gIsFeatPopnOL", >>> + >>> BGKmd >>> ="gBGMedianSignal"), >>> + columns=list(Rf="gTotalGeneSignal", >>> + >>> Gf="gTotalProbeSignal", >>> + >>> Rb="gTotalGeneSignal", >>> + >>> Gb="gProcessedSignal"), >>> + verbose=TRUE,sep="\t",quote="") >>>> ddNORM = tgsNormalization(ddTGS, "quantile", makePLOTpre = T, >>> makePLOTpost = T, targets.micro, verbose = TRUE) >>> Error in density.default(object[, n], na.rm = TRUE) : >>> need at least 2 points to select a bandwidth automatically >>>> >>>> ddNORM = tgsNormalization(ddTGS, "quantile", makePLOTpre = F, >>> makePLOTpost = F, targets.micro, verbose = TRUE) >>> Error in xy.coords(x, y) : 'x' and 'y' lengths differ >>>> >>>> >>>> ddTGS.rma = rmaMicroRna(ddaux, normalize = TRUE, background = TRUE) >>> Error in split.default(0:(length(pNList) - 1), pNList) : >>> Group length is 0 but data length > 0 >>> # this takes quite a few minutes to process, then gives this error >>> >>> I've seen quite a bit of Agilent microRNA data through our centre, >>> and >>> can't recall ever seeing a chr_coord column, so is this to do with >>> different versions of Agilent Feature Extraction, or different >>> defaults set by the array facility? >>> >>> I'd really like to RMA normalize these data, so any help would be >>> really appreciated >>> >>> cheers, >>> Mark >>> >>> >>> sessionInfo() >>> R version 2.11.1 (2010-05-31) >>> i386-apple-darwin9.8.0 >>> >>> locale: >>> [1] en_AU.UTF-8/en_AU.UTF-8/C/C/en_AU.UTF-8/en_AU.UTF-8 >>> >>> attached base packages: >>> [1] stats graphics grDevices utils datasets methods base >>> >>> other attached packages: >>> [1] AgiMicroRna_1.2.0 preprocessCore_1.10.0 affy_1.26.1 >>> limma_3.4.3 Biobase_2.8.0 >>> >>> loaded via a namespace (and not attached): >>> [1] affyio_1.16.0 tools_2.11.1 >>>> >>> >>> >>> ----------------------------------------------------- >>> Mark Cowley, PhD >>> >>> Peter Wills Bioinformatics Centre >>> Garvan Institute of Medical Research, Sydney, Australia >>> ----------------------------------------------------- >>> >>> >>> [[alternative HTML version deleted]] >>> >>> _______________________________________________ >>> Bioconductor mailing list >>> Bioconductor at stat.math.ethz.ch >>> https://stat.ethz.ch/mailman/listinfo/bioconductor >>> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor >> >> _______________________________________________ >> Bioconductor mailing list >> Bioconductor at stat.math.ethz.ch >> https://stat.ethz.ch/mailman/listinfo/bioconductor >> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor