Reference-guided transcriptome assembly using multiple SRA experiments ?

Question

[debitboro]

Dear Biostars,

I want to perform a genome-guided transcriptome assembly and would like to know whether I can use RNA-seq libraries from different NCBI SRA projects related to the same organism and strain. I want to collect several RNA-seq libraries from different conditions for each project, perform the assembly for each project, and then combine them to create a complete transcriptome. I have no idea of the extent to which the batch effect may impact the final results, as the experiments were performed separately. Any idea?

[lieven.sterck]

Sounds like a sensible approach to me.

I'm sure that there must be papers around already that have done more or less the same, no? I'm thinking towards pan-transcriptome analyses for instance (not exactly the same but good enough to give you some valuable ideas and/or pointers)

If your goal is to get as much as possible transcripts than batch effect will not be an issue I assume. Moreover technically you actually want as much batch effect as possible since that will result in many condition/experiment-specific transcripts ;-)

If later-on you want to perform more downstream analyses with this data you might need to reconsider this, but simply for the goal of building the transcriptome I would not worry much .