How to generate a reference with several tissues?

Question

[jyshiningsoul]

Hi, in my experiment design, I have RNA-seq data of three different tissues, and I want to take a look at the expression pattern of three tissues. But I have a question: how to generate a reference for my experiment design? Should I put the reads of three tissues together to get a assembly or generate three assemblies according to different tissues using Trinity? Thank you.

[WouterDeCoster]

You didn't specify this, but I assume that your species does not have a reference genome?

[jyshiningsoul]

Yes, the specie I am studying is a non-model organism, and now I have RNA-seq data of its three tissues.

Answer 1

[lakhujanivijay]

If you have the reference genome available for the organism from which the tissues were extracted , you can go for a what is known as a "reference" based transcriptome assembly. If you don't have a reference available as in case of certain non-model / less studied/ less known / less popular organisms, you can perform what is known as a "denovo master" assembly using all the 3 samples together.

Extra piece of information

Since a genome contains the sum of all introns and exons that may be present in a transcript, spliced variants that do not align continuously along the genome may be discounted as actual protein isoforms. Even if a reference genome is available, de novo assembly should be performed, as it can recover transcripts that are transcribed from segments of the genome that are missing from the genome assembly

Source: https://en.wikipedia.org/wiki/De_novo_transcriptome_assembly

[jyshiningsoul]

Thank you for your reply. Actually, I am studying on a non-model organism. I have generated two assemblies to test which assembly is better: (1) assembly1: I pooled samples of three tissues together, (2)assembly2,3,4: reads from three tissues were generated according to tissues, then I got three assemblies, then the reads of three tissues were mapped to respective references. Then I followed the Trinity pipeline, I got the count matrix for next DEG analysis. The problem is the expression patterns of three tissues based on assembly1 or assembly2,3,4 are different, so I am wondering which assembly plan is better?