Hello,

apologies, I am confused by the following. It seems that people often use Deseq2 to identify differential amplitude of gene expression, but dexseq to identify alternative splicing. 

However, Deseq2 can kindof give alternative splicing information if analysis is done at transcript level rather than gene level. 

The species I work with (ants and bees) have almost no preexisting information about alternate transcripts. Such information is thus limited or incorrect. 

Is it possible to tell deseq2 to do an exon-level analysis? (I guess that this may require exon-specific counts as input)? Is there any reason not to do this?  The potential benefit for me in using Deseq2 for this is that I would not have to change the majority of my analysis toolset.

(we are using a pseudoaligner, kallisto, and suppose that we would have to give it a slightly modified GFF as input)

Thanks v much in advance,

Yannick

You should certainly use DEXSeq if you don't know much about the transcripts. It is powerful for detecting differential exon usage, and doesn't require knowledge of which exons are combined into transcripts. DEXSeq requires that you align the reads to the genome, but the latest batch of read aligners are very fast. See HISAT2, for example, which has very low memory requirement (< 5 GB) and can align >100,000 reads a second.