Hi Simon, I have a distantly related question -- suppose we wanted to compute FPM, i.e. fragments per million, for paired-end RNAseq data. Like, say, if a person (me) was really really lazy and was less concerned about sensitivity, since they were just going to feed everything to voom() or fit robust linear models of log(RPM/FPM) against predictive factors. Is this (fragment-level computation) possible in an existing Bioconductor package at this point in time? Or using HTseq, for that matter? The nice thing about FPM or log1p(FPM) or the like is that it makes sense for the type of modeling that a lot of us do, a lot of the time. Thanks for any information and/or thoughts you might have on the topic, and for making DEXSeq available, --t On Fri, Sep 7, 2012 at 2:20 PM, Simon Anders wrote: > Dear Andres > > > On 09/07/2012 05:05 PM, Andres Eduardo Rodriguez Cubillos wrote: > >> Overall, I organized my tables quite similar to the >> pasilla_gene_counts.tsv file you used in your guide "Analyzing >> RNA-seq Data with the DESeq Package" from 2012. At first I had two >> tables: one per replicate containing FPKMs for the three conditions >> analyzed. I then merged the FPKMs of both files to obtain three >> > > > FPKM values are not suitable as input data for DESeq. Please see the > vignette, which states, in Section 1: > > The count values must be raw counts of sequencing reads. This is > important for DESeq's statistical model to hold, as only raw reads allow > to assess the measurement precision correctly. (Hence, do not supply > rounded values of normalized counts, or counts of covered base pairs.) > > > Simon > > ______________________________**_________________ > 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 > -- *A model is a lie that helps you see the truth.* * * Howard Skipper [[alternative HTML version deleted]]