Hi, I have been working with edgeR to find differentially expressed genes for RNA-seq data. I have been working with a data set with 3 treatment groups and a total of 10 samples per treatment group. The samples were sequenced as single-end, stranded reads. I first analyzed this dataset with the edgeR v2.6 and was getting 100-300 ( FDR<0.05, tagwise dispersion with prior.n=20) differentially expressed genes for each pairwise comparison. I upgraded to version 3.2.4 this weekend and reanalyzed the same dataset. I now get <100 genes as being differentially expressed (FDR<0.05, tagwise dispersion with prior.df=20) across comparisons. Does anyone know why there would be such a big difference in # of genes being called DEGS? The smaller gene list is complete subset of the larger gene list so I am assuming that some upgrades caused edgeR to be more conservative. Thanks, Marsha -- output of sessionInfo(): R version 3.0.1 (2013-05-16) Platform: x86_64-apple-darwin10.8.0 (64-bit) locale: [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8 attached base packages: [1] stats graphics grDevices utils datasets methods base other attached packages: [1] edgeR_3.2.4 limma_3.16.7 -- Sent via the guest posting facility at bioconductor.org.

Dear Marsha, This comes up at least every month or so. The short answer is that prior.n and prior.df are not the same thing, so setting them both to 20 is using very different settings, which is why you're getting different results. I believe the edgeR documentation has information on the conversion, in case you want to convert your old setting of prior.n=20 to prior.df. You can also search the archives of this list, where this question has been answered several times. Regards, -Ryan Thompson On Tue 13 Aug 2013 10:06:22 AM PDT, mwheeler [guest] wrote: > > Hi, > > I have been working with edgeR to find differentially expressed genes for RNA-seq data. I have been working with a data set with 3 treatment groups and a total of 10 samples per treatment group. The samples were sequenced as single-end, stranded reads. I first analyzed this dataset with the edgeR v2.6 and was getting 100-300 ( FDR<0.05, tagwise dispersion with prior.n=20) differentially expressed genes for each pairwise comparison. I upgraded to version 3.2.4 this weekend and reanalyzed the same dataset. I now get <100 genes as being differentially expressed (FDR<0.05, tagwise dispersion with prior.df=20) across comparisons. Does anyone know why there would be such a big difference in # of genes being called DEGS? The smaller gene list is complete subset of the larger gene list so I am assuming that some upgrades caused edgeR to be more conservative. > > Thanks, > Marsha > > -- output of sessionInfo(): > > R version 3.0.1 (2013-05-16) > Platform: x86_64-apple-darwin10.8.0 (64-bit) > > locale: > [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8 > > attached base packages: > [1] stats graphics grDevices utils datasets methods base > > other attached packages: > [1] edgeR_3.2.4 limma_3.16.7 > > > -- > 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