Kallisto Bootstrapping and deseq2

Question

[Mozart]

Dear all, Just a quick clarification about importing pseudocounts into the deseq2 package. I don't usually apply the bootstrapping step with the Kallisto package for deseq2 analysis. My results are definitely comparable to ones generated by technicians in my university with totally different approaches. Nevertheless, my question is whether it could be considered a mistake not to run the pseudoalignment step without the bootstrap part when the tsv files are then processed by the tximport package for the downstream analysis with deseq2. Thanks

Answer 1

[i.sudbery]

DESeq2 has no way to take the bootstrap analysis from Kalisto (or Salmon) into account when doing DE. Bootstraps can only be used by tools designed to use them. As far as I'm aware, at the moment, the only DE tools that can do this are the swish method from Fishpond and the sleuth package.

[Mozart]

Thanks for confirming this. This clarify things up!