Human De Novo Assembly from RNA-Seq data?

Question

[wangfx]

Hi all,

I was wondering what the best way to assemble a genome transcriptome(sorry, I typed down genome instead of transcriptome) from human RNA-seq data would be; I've looked at Trinity, AByss and the like, but I am a bit lost in the process of choosing between these software. It seems like these programs work far better with smaller genomes, so I was wondering if there were any more human oriented assembling programs. Additionally, I was wondering whether one would use multiple sets of RNAseq data to aid in assembling a transcriptome (the same sample/cell type has been sequenced multiple times) . I have not seen anything about this yet, but that may also be because I haven't looked hard enough.

Thanks for reading this!

Edit:

I should probably add that the reads are generally over 100 million in count and average around 50bp per read.

Answer 1

[h.mon]

You will never assemble a genome from RNAseq data, you will be able to assemble a transcriptome though. However, for human data, you do not need to assemble the transcriptome, mapping to the reference will provide better results using fewer resources for most analyses.

edit: anyway, yes, you can pool all your samples to assemble a "reference transcriptome". Trinity is capable of assembling human data, depend on your computational resources, especially memory available.