Hi list, Sorry to bother you but I do have a problem when testing the codes in the vigenette of *lumi* using my own data. As of my understanding, *lumi* vignette presents two methods of processing Illumina raw data, one is to process in a step-wise manner and the other one is called 'lumiExpresso' to "encapsulate" the processing steps. But the underlying mechanism should remain the same (correct me if I am wrong). I tried both of these two methods and append the codes at the end of message. However, I found the difference between the generated expression values is huge. Using the expression values generated from the first method (step- wise), quite a few differentially expressed genes could be identified afterwards. For the data generated from ''lumiExpresso', the data seems to me of log 2 transformation or somewhat similar (I can send the expression data if you need for reference). I am anxious to seek your help on where I perhaps got messed up. Thank you so much for your suggestions. Sincerely, Allen ************************1st code: step-wise manner************************ ***** library(lumi) library(EBarrays) fileName <- 'Susan_RawData_0515082.txt' example.lumi <- lumiR(fileName, lib='lumiHumanV2',columnNameGrepPattern = list(exprs='AVG_Signal', se.exprs='BEAD_STDEV', detection='Detection', beadNum='Avg_NBEADS')) summary(example.lumi, 'QC') lumi.B <- lumiB(example.lumi, method='bgAdjust') lumi.T <- lumiT(example.lumi) lumi.N <- lumiN(lumi.B, method='rsn', ifPlot=TRUE) lumi.N.Q <- lumiQ(lumi.N) summary(lumi.N.Q, 'QC') write.exprs(lumi.N.Q, file='processedSusanData0530NQ.txt') ************************2nd code: using 'lumiExpresso'***************************** fileName <- 'Susan_RawData_0515082.txt' example.lumi <- lumiR(fileName, lib='lumiHumanV2',columnNameGrepPattern = list(exprs='AVG_Signal', se.exprs='BEAD_STDEV', detection='Detection', beadNum='Avg_NBEADS')) summary(example.lumi, 'QC') lumi.N.Q <- lumiExpresso(example.lumi, QC.evaluation=TRUE) summary(lumi.N.Q, 'QC') write.exprs(lumi.N.Q, file='processedExampleData.txt') [[alternative HTML version deleted]]

On Fri, May 30, 2008 at 1:10 PM, ss <affysnp at="" gmail.com=""> wrote: > Hi list, > > Sorry to bother you but I do have a problem when testing the codes in > the vigenette of *lumi* using my own data. > > As of my understanding, *lumi* vignette presents two methods of processing > Illumina raw data, one is to process in a step-wise manner and the other one > is called 'lumiExpresso' to "encapsulate" the processing steps. But the > underlying > mechanism should remain the same (correct me if I am wrong). > > I tried both of these two methods and append the codes at the end of > message. > However, I found the difference between the generated expression values is > huge. > Using the expression values generated from the first method (step- wise), > quite > a few differentially expressed genes could be identified afterwards. For the > data > generated from ''lumiExpresso', the data seems to me of log 2 transformation > or > somewhat similar (I can send the expression data if you need for reference). > > > I am anxious to seek your help on where I perhaps got messed up. Thank you > so much for your suggestions. Hi, Allen. You will notice if you read the help for lumiExpresso (you really HAVE to read the help--the vignette is not enough), you will notice that the function accepts a list of parameters for each of the lumiB, lumiT, and lumiN steps. You will want to carefully match the arguments for each of those if you want to compare results. I did a quick test and got the same results using both methods. If you match all the parameters and still get different results, then perhaps you can post a reproducible example. Also, remember to ALWAYS include sessionInfo() when writing to the list. Sean > > ************************1st code: step-wise manner************************ > ***** > library(lumi) > library(EBarrays) > fileName <- 'Susan_RawData_0515082.txt' > example.lumi <- lumiR(fileName, lib='lumiHumanV2',columnNameGrepPattern = > list(exprs='AVG_Signal', se.exprs='BEAD_STDEV', detection='Detection', > beadNum='Avg_NBEADS')) > summary(example.lumi, 'QC') > lumi.B <- lumiB(example.lumi, method='bgAdjust') > lumi.T <- lumiT(example.lumi) > lumi.N <- lumiN(lumi.B, method='rsn', ifPlot=TRUE) > lumi.N.Q <- lumiQ(lumi.N) > summary(lumi.N.Q, 'QC') > write.exprs(lumi.N.Q, file='processedSusanData0530NQ.txt') > > > ************************2nd code: using > 'lumiExpresso'***************************** > fileName <- 'Susan_RawData_0515082.txt' > example.lumi <- lumiR(fileName, lib='lumiHumanV2',columnNameGrepPattern = > list(exprs='AVG_Signal', se.exprs='BEAD_STDEV', detection='Detection', > beadNum='Avg_NBEADS')) > summary(example.lumi, 'QC') > lumi.N.Q <- lumiExpresso(example.lumi, QC.evaluation=TRUE) > summary(lumi.N.Q, 'QC') > write.exprs(lumi.N.Q, file='processedExampleData.txt') > > [[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 >