Should RNA-seq data be assembled before analyzing?

Question

[nancydong20]

Hello!

Should RNA-seq data (almost) always be assembled first into contigs before analyzing? From what I can see, not all published transcriptome studies assemble the reads before proceeding onto expression level analysis.

My understanding is that assembling into contigs will enable better ORF/gene detection, but are there analyses that can be done before assembling?

Thank you very much for your inputs!

Answer 1

[GenoMax]

Only if you don't have a reference genome available. Otherwise align to genome.

[nancydong20]

Thank you very much for your reply!

There is a draft genome for this organism, but it's only on its first release. I worry that by aligning the reads to a maybe not-so-accurate genome can cause me to lose out on useful information in the transcriptome?

[GenoMax]

Do both. Even with the draft genome you would get an idea of what your data looks like projected against the genome. If this is a eukaryotic genome then you are likely to use Trinity to assemble the transcriptome. This could require significant amount of compute resources.