RNA-sequenceing mode for variant calling - PE or SE

Question

[Alex V. Gopanenko]

Dear community,

I wonder if there are any guidelines/best practices to select the RNA-seq sequencing mode (I mean SE or PE and also the length.. 75, 100, and so on) for variant calling?

My case: I have paired samples of total RNA (one - treated and has characteristic mutation patterns (conversions), and one - mock). So, I want to discover all treatment-dependent conversion events. I understand that PE mode outperforms SE in terms of mapping quality and so on, but SE also is cheaper.

Could you please provide any consideration why to use SE or PE sequencing? What should be the length of reads and so on?

I will appreciate any suggestions,

Thanks in advance,

Best regards,
Alex

Answer 1

[ATpoint]

RNA-seq is in no way optimal for variant calling (reverse transcription, many copies of a gene as mRNA molecules on top of clonal heterogeneity, lots of PCR cycles, lots of duplicates...) so at least do paired-end with reads as long as possible to have good coverage. 2x150bp is standard these days.

That having said, WGS or targeted sequencing would have been the better choice here.

[Alex V. Gopanenko]

Thanks a lot for your reply! Yes, I understand, that WGS/WES is better in terms of variant calling. I'm working on the identification of RNA modification, so after the chemical treatment of RNA samples I have conversion events that I want to identify. The only choice for me is RNA-seq. Anyway, thanks for the suggestion to use 150 PE.

[ATpoint]

I see. 2x150 will maximize coverage and by this maximize information to work with.