Thank you. Yes, this was the case (I am new with R), so I was unaware of the implementation in sva. Lilach On Sat, Dec 24, 2011 at 8:45 PM, Dan Tenenbaum <dtenenba@fhcrc.org> wrote: > On Sat, Dec 24, 2011 at 6:09 AM, Lilach Traivish > <lilachtraivish@gmail.com> wrote: > > Dear list members, > > > > I applied ComBat to control for batch effect in Affymetrix microarray > data. > > > > I managed to get results from ComBat only when adding the skip parameter > as > > follows: > > ComBat('exp_data.txt','sample_information_file.txt',skip=1); > > > > However, the header row of the results file misses one value - at the end > > of the header row. > > The missing sample same is actually of the first sample which is present > in > > the sample information file. Thus, I assume that all the sample names are > > given in an offset of 1. Is is possible that this is the case? As I wish > to > > use the output of ComBat as an input for differential expression > analysis, > > how can this issue be resolved? > > It looks like you are not asking about the ComBat method in the sva > package, but instead a script that is not a part of Bioconductor. > > You should probably contact the maintainer of that script > (cplaisier@systemsbiology.org). > > Dan > > > > > Thanks in advance > > > > Regards, > > Lilach > > > > [[alternative HTML version deleted]] > > > > _______________________________________________ > > Bioconductor mailing list > > Bioconductor@r-project.org > > https://stat.ethz.ch/mailman/listinfo/bioconductor > > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]

Thank you again. Following your previous answer I have tried to use the sva package. I have read the documentation PDF file. I have run the following code on the data - > biocLite("sva") > biocLite("affy") > library(affy) > RMAdata<-justRMA(celfile.path=dataDir) > pheno = read.phenoData("pheno_data.txt", header = TRUE, row.names = 1) My pheno data file contains 6 columns and a header line, the first two columns are array and sample id numbers and the rest are the variables to adjust for. Now I wish to describe the model to be fit by creating a model matrix. In the sva documentation file code example, it is unclear to me, what is the 'cancer' variable > mod = model.matrix(~as.factor(cancer), data=pheno) is it a vector with no header and the sample status? (I did not manage to get bladderbatch library through R). If I try to create a model matrix using only the batch (scan date) variable > mod=model.matrix(~as.factor(batch),data=pheno) I received the following error: Error in as.data.frame.default(data, optional = TRUE): cannot coerce class 'structure("AnnotatedDataFrame", package = "Biobase")' into a data.frame Is there a way to consider more than 1 variable known batch correction (e.g scan date, gender) in describing the model matrix? If there is another documentation source to refer to regarding this issue I will appreciate it (as this seems like a basic question). My R version: ------------------------------------------------------ > version _ platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major 2 minor 14.0 year 2011 month 10 day 31 svn rev 57496 language R version.string R version 2.14.0 (2011-10-31) Thank you! Lilach On Sat, Dec 24, 2011 at 8:45 PM, Dan Tenenbaum <dtenenba@fhcrc.org> wrote: > On Sat, Dec 24, 2011 at 6:09 AM, Lilach Traivish > <lilachtraivish@gmail.com> wrote: > > Dear list members, > > > > I applied ComBat to control for batch effect in Affymetrix microarray > data. > > > > I managed to get results from ComBat only when adding the skip parameter > as > > follows: > > ComBat('exp_data.txt','sample_information_file.txt',skip=1); > > > > However, the header row of the results file misses one value - at the end > > of the header row. > > The missing sample same is actually of the first sample which is present > in > > the sample information file. Thus, I assume that all the sample names are > > given in an offset of 1. Is is possible that this is the case? As I wish > to > > use the output of ComBat as an input for differential expression > analysis, > > how can this issue be resolved? > > It looks like you are not asking about the ComBat method in the sva > package, but instead a script that is not a part of Bioconductor. > > You should probably contact the maintainer of that script > (cplaisier@systemsbiology.org). > > Dan > > > > > Thanks in advance > > > > Regards, > > Lilach > > > > [[alternative HTML version deleted]] > > > > _______________________________________________ > > Bioconductor mailing list > > Bioconductor@r-project.org > > https://stat.ethz.ch/mailman/listinfo/bioconductor > > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]