I've done standard rna-seq differential expression analysis on a gene level using edgeR. For downstream analysis I would like to extract the normalized counts for a particular exon. What is the best way to go about it? Can I derive the normalized counts for that exon straight from my analysis (e.g. by using, say, cpm())? The first workaround that pops to my mind is that I could simply edit either the GTF or the raw count data to discard any counts for other exons of that gene, than simply run the standard count normalisation and I would get the normalized counts for that exon masquerading as the normalized count for the corresponding gene. But this seems too hacky and dirty solution. Can anyone give advice on how to do this properly? [[alternative HTML version deleted]]

There is no way to derive exon-level counts from gene-level counts. You will have to go back to the bam files and perform a separate counting step for exons. On Sat Mar 29 17:48:02 2014, Nick N wrote: > I've done standard rna-seq differential expression analysis on a gene level > using edgeR. For downstream analysis I would like to extract the normalized > counts for a particular exon. What is the best way to go about it? Can I > derive the normalized counts for that exon straight from my analysis (e.g. > by using, say, cpm())? > > The first workaround that pops to my mind is that I could simply edit > either the GTF or the raw count data to discard any counts for other exons > of that gene, than simply run the standard count normalisation and I would > get the normalized counts for that exon masquerading as the normalized > count for the corresponding gene. But this seems too hacky and dirty > solution. Can anyone give advice on how to do this properly? > > [[alternative HTML version deleted]] > > _______________________________________________ > 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

I don't think you can get exon-level counts from a gene-level analysis because your read counts were already summarized for genes. To get normalized counts for exons, you need to get raw counts for exons first. To do this, you can use the featureCounts command in Rsubread package. library(Rsubread) ?featureCounts featureCounts accepts GTF annotation as input. You should set useMetaFeatures=FALSE and allowMultiOverlap=TRUE when you try to get read counts for exons. Note that featureCounts can get you read counts for genes as well. Hope this helps. Wei On Mar 30, 2014, at 11:48 AM, Nick N wrote: > I've done standard rna-seq differential expression analysis on a gene level > using edgeR. For downstream analysis I would like to extract the normalized > counts for a particular exon. What is the best way to go about it? Can I > derive the normalized counts for that exon straight from my analysis (e.g. > by using, say, cpm())? > > The first workaround that pops to my mind is that I could simply edit > either the GTF or the raw count data to discard any counts for other exons > of that gene, than simply run the standard count normalisation and I would > get the normalized counts for that exon masquerading as the normalized > count for the corresponding gene. But this seems too hacky and dirty > solution. Can anyone give advice on how to do this properly? > > [[alternative HTML version deleted]] > > _______________________________________________ > 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 ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:6}}