Analyzing RNA Seq splice variants using spike-ins?

Question

[jc.szamosi]

I'm about to start a splice variant analysis in RNASeq, and I was wondering if anyone has done this using spike-in controls.

I see that ERCC normalization can't be combined with CuffDiff2, because CuffDiff2 has its own normalization, but I was wondering if anyone is using ERCC controls anyway just as a sanity check, and to look at things like minimum limit of detection.

I've also found SIRV spike ins that are specifically intended for use with splice variant analysis, and I was wondering if anyone has any experience with them, and if CuffDiff2 is an appropriate tool to use in that case.

Thanks!

[Satyajeet Khare]

I was under an impression that type of normalization (spike in or no spike in) won't make much difference to the detection of alternative splicing events, since it is relative expression analysis of isoforms of the parent gene in two samples. Even the lab validation of relative isoform expression is generally done by normalization to the expression to the parent gene. Let know your experiences.

Answer 1

[nwon]

Bump!

I am also interested to see experiences of other colleagues, ERCC is good for LOD, but also reveals and confirms what is known, it's not a sensitive as RT-qPCR.

Another spike in that just came live are sequins, there is a Nature Comm paper on this and a website to gain access to them too! Link to Sequins