Is ddRAD sequencing strand-specific?

[carladosanjos]

Hi all,

I have ddRAD libraries that I am trying to build an assembly-based reference, but I am not sure if ddRADseq sequencing reads are strand-specific. Are they? I cannot find this information anywhere! Upon assembling my libraries, the data isn't looking strand specific.

Does anybody knows the answer/

Many thanks!

[nucacidhunter]

Yes they are. Library is from bottom strand.

[carladosanjos]

Many thanks!
Could you be more specific? I really want to understand how it works. Why do I have forward/reverse reads perfectly aligned? Does it means they likely are from different genome regions?

[SNPsaurus]

Which enzymes did you use? Is it paired-end sequencing? What fragment size?

If you allow short fragments and did paired-end sequencing, then you might get overlapping reads and you'll see this for short fragments but not longer ones. Perfectly aligned doesn't sound like overlapping to different degrees, though. Are the reads aligning all the way from the first to last base? With your alignment parameters would you see partial overlaps?

I think ddRAD has some artifacts from exonucleases acting on overhangs, allowing newly-blunt adapters to ligate to random fragments. You would see some sporadic, random forward and reverse reads from that, but low level. Give us an example of this with how many forward and how many reverse and the sequences, that might yield a clue.

[nucacidhunter]

Many thanks!
Could you be more specific? I really want to understand how it works. Why do I have forward/reverse reads perfectly aligned? Does it means they likely are from different genome regions?

As shown in attached diagram from ddRAD paper https://docs.google.com/document/d/1.../edit?hl=en_US
priming PCR starts from rare cutter terminal bottom strand (Illumina's P5 adapter) and top strand will not be amplified and does not have complete P5 and P7 adapters necessary for sequencing.