Based on the previous threads when using DESeq the reads should not be double counted. I am following pipeline for RNA-seq analysis and would like to know any suggestions/comments regarding the pipeline: 1. Mapping the reads using Tophat 2. Convert Tophat output.bam to Sam 3. Create bed file from Sam file 4. Use CoverageBed along with reference genome for counting the reads 5. Sum count of reads from all exons in a transcript 6. DESeq to analyze the counts/transcript Thank you very much Nirmala [[alternative HTML version deleted]]

Unless you are counting reads on the unique disjoin set of exons from the transcripts , double counting will be inevitable as most of the transcripts for a gene will have overlapping exons. You can also see Simon's post today on a different thread about the same issue. He explains in more detail why one would want to stay with gene level or exon level differential expression and relate it back to gene/isoforms. -Abhi On Thu, Feb 9, 2012 at 9:06 AM, Akula, Nirmala (NIH/NIMH) [C] <akulan at="" mail.nih.gov=""> wrote: > Based on the previous threads when using DESeq the reads should not be double counted. I am following pipeline for RNA-seq analysis and would like to know any suggestions/comments regarding the pipeline: > > > ?1. ?Mapping the reads using Tophat > ?2. ?Convert Tophat output.bam to Sam > ?3. ?Create bed file from Sam file > ?4. ?Use CoverageBed along with reference genome for counting the reads > ?5. ?Sum count of reads from all exons in a transcript > ?6. ?DESeq to analyze the counts/transcript > > Thank you very much > Nirmala > > > > > ? ? ? ?[[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