mRNA sequencing - stranded versus not

[N311V]

Hi all,

We're in the early stages of planning an RNAseq experiment with two goals. One is eQTL and or differential expression analysis and the other is denovo assembly of the transcriptome. Library prep of the two goals does not necessarily have to be the same.

My question is, would you advice for or againsts using stranded versus unstranded library prep for these goals?

Thanks

[blancha]

We only do stranded RNA-Seq at our institute.

It eliminates the ambiguity when you have genes at the same location on opposite strands. This is often the case for non-coding RNA, and protein-coding genes. The results can be distorted if you assign the reads from a non-coding RNA to a protein-coding gene.

I only see advantages to stranded RNA-Seq which is why we do it systematically, regardless of the objective of the experiment.

I don't do the library prep myself, only the bioinformatics analysis. The molecular biology platform at our institute does the library prep using a dUTP protocol. Since they do it systematically, partly upon my recommendation, I can't imagine it adds much work to the library prep.

[N311V]

We only do stranded RNA-Seq at our institute.

It eliminates the ambiguity when you have genes at the same location on opposite strands. This is often the case for non-coding RNA, and protein-coding genes. The results can be distorted if you assign the reads from a non-coding RNA to a protein-coding gene.

I only see advantages to stranded RNA-Seq which is why we do it systematically, regardless of the objective of the experiment.

I don't do the library prep myself, only the bioinformatics analysis. The molecular biology platform at our institute does the library prep using a dUTP protocol. Since they do it systematically, partly upon my recommendation, I can't imagine it adds much work to the library prep.

Thanks for the reply, it was very helpful. I don't know how often genes will overlap on opposite strands for the species I'm working with but I'd be surprised if it doesn't happen a fair bit. So far based on this insight I'd say stranded is definitely the way to go.

I'm not sure if it's the case or not but the only down side I envisages would be if more reads per sample are needed for stranded vs non-stranded RNA seq.