Dada2 Single or paired?


I am new to QIIME2 and after some tutorials I decided to analyse my won data so:
I have raw sequencing data of paired reads, and I have run them with dada2 denoise-single and results look good.
Now the question is, do I need to run denoise-paired on paired reads data or is it just my choice in a way how i want to analyse my samples (single or paired approach)?

Best regards,

Welcome to the forum!
In your current analyses, only forward reads were used, and reverse reads were discarded since you run Dada2 for single reads. If you will rerun Dada2 for paired reads, it will merge forward and reverse reads to longer reads, wich will improve taxonomy assignment and feature resolution. Usually reverse reads are discarded only if the quality is too low or reads are failing to merge.


Thank you!

Thats literally the case. I have reruned with Dada2 paired reads and lost like 95-100% of samples, so in that case choosing single reads is a decent choice. Once again thank you for an explanation!

Happy New Year to all!

Did you check at which step most of the reads were lost? For example, too high truncation values can course big losses since all the reads lower than that value would be discarded at filtering step. From the other side, too low values can decrease overlapping region and a lot of reads would be discarded at merging step. So I would first play with Dada2 settings for paired reads, and use Dada2 single only in case if nothing works for paired option.

Happy New Year!

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.