I am stuck with the denoising stuff. I am not sure about the trim/trunc length. I have read multiple forum yet I couldn't decide the parameters. I am currently doing 16S V3-V4 region of MiSeq data. Any tips? Thank you.
Wow, those are some really nice quality plots! I’d say your sequencing run went very well.
Which is really good since the V3-V4 region is fairly long and can be hard to hit (depending on your specific primers I think you only have about ~90 bases free after 2x300 sequencing for that region).
Your forward reads are excellent, I would probably keep the entire length or trim at ~280 (use trunc-len of 0 to skip truncation in that direction).
You reverse is also great, I would probably trim around ~270, but you have enough wiggle room that you could trim a little closer like ~240 if you kep the full length on the forward reads.
Then see if you get good merging from your denoising stats. Ideally you won’t see much/any drop-off at that point, but let us know if the counts before/after merging don’t look good.
Oh and make sure you trim off your forward primer’s length with trim-left. That may cover the slightly lower quality on the start of the read (which is very normal), or if your lower quality is a bit longer, go ahead and extend the trim-left a few more bases.