Hi everyone,
I am exploring the possibilities I have for an RNA meta-transcriptomic sequencing experiment involving Nanopore. Disclaimer #1: I'm a bioinformatician, so I don't have much wet lab background to choose a couple of things in my workflow. Disclaimer #2: I will post this also on the Nanopore community forum, but as many of you probably know, that's not publicly accessible so I guess this justifies the crossposting here.
We have a total of ~80 meta-transcriptomic samples representing 24 conditions, each with slightly different expression profiles. So we do expect different transcripts present in the different samples, although not radically.
To perform such an experiment, would you:
- combine all samples to increase the total RNA, or do multiple different sequencing runs? (ideally one per condition)
- go through a cDNA step before sequencing, or use direct RNA sequencing?
We have abundant short reads to perform an expression analysis between conditions, so the long-reads are to have an assembled meta-transcriptome mostly.
Thanks in advance for your input!
I think you should mainly look at the difference in yield between both approaches. As far as I know they correlate reasonably well, but if I remember correctly the yield of the cDNA library prep is much higher. As an additional difference: with RNA sequencing you can identify nucleotide modifications.
True. Thanks for the point of view!
As another informatician, my comments are from a theoretical view. Direct RNA sequencing is a relatively new application for NanoPore. I am not sure how robust it is in terms of ease of use and data quality. cDNA on the other-hand should be a stable application. But if you have ample sample (and budget) then you could try the direct RNA method first.
We do have ample sample and we have enough budget. The most crucial part is to choose a path that returns sufficient results because a person in the project has limited time to complete their part of the analysis :)