Transcriptome assembly -mixing stranded and non-stranded data

Question

[Lada]

hello. I am new in Bioinformatics so I am wondering is it possible to make transcriptome assembly using combined sequences: stranded/directional and nonstranded?

I am making de novo transcriptome assembly for one speciese without a reference genome and I have the stranded/directional seq data 20M (RNA from whole individuals) and non-stranded/standard seq data with 20M and 10M (RNA from particular tissue etc.). Both 2x150bp.

So i was thinking to use all my sequences in order to make "better" transcriptome coverage but I don't know if it is possible to combine those two type of data? Any help will be very appreciated ☺️

Answer 1

[lieven.sterck]

I would personally prefer the stranded data, not only because it's stranded but also because it's whole individual.

technically there is likely very little issue combining them, so you might give that a try indeed. Alternatively you can do an assembly with each dataset and then combine the results afterwards to get to a more complete transcriptome.

[Lada]

Hi, thanks for your answer. :)

The reason why I wanted to include the non-stranded data from a specific tissue in the transcriptome assembly is that I am also interested in the gene expression within that particular tissue and these transcripts might be underrepresented when analysing the whole individual (just assuming). I'll definitely try to combine all the sequences and see how it goes.

Also, I already have 2 transcriptome assemblies - 1 from stranded and 1 from nonstranded data. So you are saying I can combine those 2 transcriptomes? how do I do that, do you maybe know? :)

[lieven.sterck]

ah, excellent :-)

goh, there are many tools around that can do this (look for "remove redundancy from assembly" or such)

one tool is for instance CDHIT-est (actually used to cluster similar sequences) from which you should be able to deduct a near non-redundant set

[Lada]

ok, tnx! :)