I was wondering, whether it make sense to run DESeq on the transcript level. I am interested on differentially expressed exons (differential exon usage) and would like to know how exons are being expressed under different conditions.
I have counted the reads with featureCounts on the transcript level:
featureCounts -a ~/genomes/Drosophila_melanogaster/genes.gtf -f -O -t exon -g transcript_name -o featureCounts.out.txt ../bamFiles/sample1.bam
this gives me the list of the exons with the number of reads for each exon.
I would like than to run DESeq2 with the count table of the single exon. But before I am doing that, I was wondering if it make sense at all to run DESeq with the exons, as the distribution of the reads over exon is not the same as for the genes. In featureCounts I also allow a read to be counted in multiple transcripts, if an exon is expressed in more than just one transcript.