Transcriptome assembly: single individual or pooled samples

Question

[mariannapauletto]

Dear colleagues,

I'm going to set up a RNA-seq experiment to characterize the transcriptome of different tissues of a species with a published genome in ncbi (but not a model species...). The aim is to characterize the tissue-specific transcriptome, evaluating eventual alternative splicing events. I'm going to make a 150 PE seq with a good coverage (at least 50M reads). The idea is to start from 3 pooled samples for each tissue in order to represent also variability occurring in different animals (what if a gene is not expressed in an individual for its specific conditions/characteristics?).

Going to the question: based on your experience, do you think that state-of-the-art transcriptome assembly software might have "problems" with reads coming from different animals? I'm thinking about polymorphisms... This should not be a problem, having a reference genome, isn't it?

Thank you in advance for your time and suggestions

Best
Marianna

Answer 1

[JC]

I don't think you will see such errors, I commonly use Trinity or rnaSPAdes to assemble pooled transcriptomes from different organisms (animals, plants, fungi), in general, short polymorphisms are not problem for modern assemblers, as they will choose a nucleotide with the highest frequency for the consensus sequence.