Hello, Here is an update of my problem :
Maybe I was not so clear with my general description: I have an exon tiling array as input (CDF) and did the first 2 steps of the RMA analysis using 6 CEL files which contains the raw intensities (3 flower replicates and 3 leaf replicates). I want to find alternative splice junctions between the flower and leaf samples. For this I want to use FIRMA. As I understood it, FIRMA uses transcript-level and exon-level data to calculate FIRMA scores, which give information e.g. on exon skipping (very low/negative log value). Since I am not able to use the the method 'setMergeGroups', here is what I tried without succeeding: 1) does a fit without group structure, csN is the quantile normalized cel set plmTr <- ExonRmaPlm(csN, mergeGroups=TRUE) fit(plmTr, unit=NULL, verbose=verbose, force=force) # this takes a while! cesTr <- getChipEffectSet(plmTr) 2) extracted the data frame to be able to set the plmTr$groupName and plmTr$group variables explicitely dfTranscript <- extractDataFrame(cesTr, units=NULL, addNames=TRUE) dfTranscript$groupName = transcriptClusterNames # set manually gene model names (transcripts), which define groups dfTranscript$group = transcriptClusterNumbers # set manually group numbering (numeric), starting from 1 to 30 080 I have now a data.frame object which gives an error when I do: > FirmaModel(dfTranscript) Error in list(`FirmaModel(dfTranscript)` = <environment>, `Arguments $getInstanceOf(rmaPlm, "ProbeLevelModel")` = <environment>, : [2011-07-11 13:54:15] Exception: Argument 'rmaPlm' is neither of nor inherits class ProbeLevelModel: data.frame at throw(Exception(...)) at throw.default(sprintf("Argument '%s' is neither of nor inherits class %s: %s", .name, class[1], paste(class(object), collapse = ", "))) at throw(sprintf("Argument '%s' is neither of nor inherits class %s: %s", .name, class[1], paste(class(object), collapse = ", "))) at method(static, ...) at Arguments$getInstanceOf(rmaPlm, "ProbeLevelModel") at FirmaModel(dfTranscript) 3) then I tried to get my data again in the 'right' data structure (should be of class ProbeLevelModel') or am I wrong? I converted it to a AffymetrixCelSet object: csTranscript = as.AffymetrixCelSet(dfTranscript) csTranscript$setCdf(cdf=cdf); Does this make sense or will be there again no group information afterwards? 4) again fitting: plmTrTest <- ExonRmaPlm(csTranscript, mergeGroups=TRUE) fit(plmTrTest, unit=NULL, verbose=verbose, force=force) # this takes a while! 5) doing FIRMA firmaTest = FirmaModel(plmTrTest) fit(firmaTest, verbose=verbose) # this takes a while! # get the firma scores fs <- getFirmaScores(firmaTest) # take the log2 of the firma scores fsScores <- log2(extractDataFrame(fs)) 6) results of these FIRMA scores are not different from the ones obtained below in my first thread. fsScores contains again specific FIRMA scores for each exon and not for each transcript cluster. I get again entries with ones (in the log scale zeros) for the fsScores $group variable. I think I have to change the internal structure of plmTr... How can I achieve this? I need the transcript cluster information for finding isoform, is'nt it? I hope I am clear in my description and would be grateful for any input on this. Thanks alot for your help! Best wishes, Marlen On 4 Jul., 16:50, marlibu <marlibu2...@gmx.net> wrote: > Hello everybody, > > first of all, thanks for providing this forum, it really helps me alot in my > data analysis. :) > > second, my problem: I have 2 costum exon CDFs with CEL files respectively. I > want to find differential splicing using FIRMA. It seems that I have after > applying the commands from the aroma.org vignette: > > plmTr <- ExonRmaPlm(csN, mergeGroups=TRUE) > fit(plmTr, unit=NULL, verbose=verbose, force=force) # this takes a while! > firma <- FirmaModel(plmTr) > #print(str(firma)) > fit(firma, verbose=verbose) # this takes a while! > # get the firma scores > fs <- getFirmaScores(firma) > #print(str(fs)) > # take the log2 of the firma scores > fsScores <- log2(extractDataFrame(fs)) > > no group information is available, I have only annotations based on exons: > > > str(fsScores) > > 'data.frame': 143259 obs. of 9 variables: > $ unit : num 0 1 1.58 2 2.32 ... > $ group : num 0 0 0 0 0 0 0 0 0 0 ... > $ cell : num 0 1 1.58 2 2.32 ... > $ caquinof_20090625_Flower1_Affy_new-v4: num 0.235 -0.183 -0.177 -0.111 > -0.094 ... > $ caquinof_20090625_Flower2_Affy_new-v4: num 0.02462 -0.21418 -0.00289 > -0.04531 0.03572 ... > $ caquinof_20090625_Flower3_Affy_new-v4: num -0.0528 -0.2674 0.0483 > -0.0398 -0.0789 ... > $ caquinof_20090625_Leaf1_Affy_new-v4 : num -0.0151 -0.1455 -0.3472 > -0.2368 -0.0481 ... > $ caquinof_20090625_Leaf2_Affy_new-v4 : num -0.207 -0.0648 -0.1754 > -0.1012 -0.4942 ... > $ caquinof_20090625_Leaf3_Affy_new-v4 : num -0.194 0.1359 0.1802 -0.4697 > 0.0343 ... > > were variables 4-6 are flower replicates and 7-9 leaf replicates of > arabidopsis plant. > But I can get the group information (exons per transcript) from the exon > annotation: > one example: > > exon annotation = 'AT1G01040.0.Chr1.plus.26862.27012' > transcript annotation = 'AT1G01040.0' > > Now, I want to 'insert' the transcript annotation as 'group' variable and I > found the method 'SetMergeGroup': > > I tried to use it like this: > > setMergeGroups(groupSet[,2]) > > but then comes the Error > > Fehler in UseMethod("setMergeGroups") : > nicht anwendbare Methode für 'setMergeGroups' auf Objekt der Klasse > "c('double', 'numeric')" angewendet > which means this: > Error in UseMethod("setMergeGroups"): > not applicable method for 'setMergeGroups' for object of the class > "c('double', > 'numeric')" applied. > > Could you tell me how to use tis method properly? Do I have some wrong > assumptions in my data processing step? > > thanks alot for your expertise in advance, > > Best wishes, > > Marlen Müller > Masterstudent at the Plant Biotechnology Group, > ETH Zurich -- When reporting problems on aroma.affymetrix, make sure 1) to run the latest version of the package, 2) to report the output of sessionInfo() and traceback(), and 3) to post a complete code example. You received this message because you are subscribed to the Google Groups "aroma.affymetrix" group with website http://www.aroma-project.org/. To post to this group, send email to aroma-affymetrix@googlegroups.com To unsubscribe and other options, go to http://www.aroma-project.org/forum/