Hi All, I am in the process of performing a DE of the RNA-seq data. I have "paired" experimental design for a "treatment".In addition to this, I have 4 categorical explanatory variables. I am aiming to look for DE genes upon "treatment" which might be dependent upon these categorical variables. The design I used: design<-model.matrix(~treat+pair+var1+var2, data=x ) When I tried to run "estimateGLMCommonDisp" using this design, I am getting the following error. Error in glmFit.default(y, design = design, dispersion = dispersion, offset = offset, : Design matrix not of full rank. The following coefficients not estimable: Do you have any idea about a correct design matrix for this kind of setting in edgeR ? Thanks a lot Nair -- output of sessionInfo(): > sessionInfo() R version 3.1.0 (2014-04-10) Platform: x86_64-pc-linux-gnu (64-bit) locale: [1] LC_CTYPE=fi_FI.UTF-8 LC_NUMERIC=C [3] LC_TIME=en_GB.UTF-8 LC_COLLATE=en_GB.UTF-8 [5] LC_MONETARY=fi_FI.UTF-8 LC_MESSAGES=en_GB.UTF-8 [7] LC_PAPER=en_GB.UTF-8 LC_NAME=C [9] LC_ADDRESS=C LC_TELEPHONE=C [11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C attached base packages: [1] grid stats graphics grDevices utils datasets methods [8] base other attached packages: [1] lattice_0.20-29 gplots_2.13.0 VennDiagram_1.6.5 edgeR_3.4.2 [5] limma_3.18.13 loaded via a namespace (and not attached): [1] bitops_1.0-6 caTools_1.16 gdata_2.13.3 gtools_3.3.1 [5] KernSmooth_2.23-12 -- Sent via the guest posting facility at bioconductor.org.

Hi Nair, You will have to give more information than that. On 4/16/2014 3:13 PM, Nair [guest] wrote: > Hi All, > > I am in the process of performing a DE of the RNA-seq data. > > I have "paired" experimental design for a "treatment".In addition to this, I have 4 categorical explanatory variables. I am aiming to look for DE genes upon "treatment" which might be dependent upon these categorical variables. The design I used: > > design<-model.matrix(~treat+pair+var1+var2, data=x ) This looks like you have only two categorical explanatory variables. It would help if you gave more information about the number of samples you have, and the different treatments. > > When I tried to run "estimateGLMCommonDisp" using this design, I am getting the following error. > > Error in glmFit.default(y, design = design, dispersion = dispersion, offset = offset, : > Design matrix not of full rank. The following coefficients not estimable: This means that you have either too many parameters for the number of samples, or one (or more) of your variables are linear combinations of the other variables. Unless you show what you are doing, nobody will be able to say much more than that. Best, Jim > > > Do you have any idea about a correct design matrix for this kind of setting in edgeR ? > > Thanks a lot > Nair > > -- output of sessionInfo(): > > > >> sessionInfo() > R version 3.1.0 (2014-04-10) > Platform: x86_64-pc-linux-gnu (64-bit) > > locale: > [1] LC_CTYPE=fi_FI.UTF-8 LC_NUMERIC=C > [3] LC_TIME=en_GB.UTF-8 LC_COLLATE=en_GB.UTF-8 > [5] LC_MONETARY=fi_FI.UTF-8 LC_MESSAGES=en_GB.UTF-8 > [7] LC_PAPER=en_GB.UTF-8 LC_NAME=C > [9] LC_ADDRESS=C LC_TELEPHONE=C > [11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C > > attached base packages: > [1] grid stats graphics grDevices utils datasets methods > [8] base > > other attached packages: > [1] lattice_0.20-29 gplots_2.13.0 VennDiagram_1.6.5 edgeR_3.4.2 > [5] limma_3.18.13 > > loaded via a namespace (and not attached): > [1] bitops_1.0-6 caTools_1.16 gdata_2.13.3 gtools_3.3.1 > [5] KernSmooth_2.23-12 > > > -- > Sent via the guest posting facility at bioconductor.org. > > _______________________________________________ > Bioconductor mailing list > Bioconductor at r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor -- James W. MacDonald, M.S. Biostatistician University of Washington Environmental and Occupational Health Sciences 4225 Roosevelt Way NE, # 100 Seattle WA 98105-6099