Hi! I have found a few similar questions but none answer this specific question:
I have data from one sequencing run on MiSeq 2x300 PE - but it was two libraries pooled together. Samples were the same (gut) but were split and processed in parallel to generate two libraries (one for the V1-V3 region, one for the V4 region). My question is, at one point in the pipeline should I treat these as separate datasets, vs. when should I keep all the sequences together? For example since everything was done in one run, is it best to keep everything together for denoising/clustering or should I split after demultiplexing and do the denoising/cluster separately?
Do the separate regions have separate barcodes, so that you can easily split them? If so, I would split them and treat them like two different data sets and never process them together ever, just like if you had sequenced 16S and ITS regions.
Dealing with multiple regions is a big challenge, but the good news is that we great thread going discussing all parts of the problem. But hopefully you can separate your regions and avoid this messiness.