By one of our first runs with MiSeq we generated data from 16S amplicon sequencing, we wonder now which data filter is more recommended using Qiime, what is your experience out there with DADA2 and Deblur?
As Ara said, there is no recommendation about which denoising algorithm to use (as far as i know). But, I wanted to mention that DADA2 is superior denoising package at present. In our lab, we use DADA2 package in R for our MiSeq data processing. You also have to consider one thing if you want to use DADA2 in Qiime2 (Qiime is discontinued), DADA2 does not produce OTUs with 97% threshold, it results into ASVs which are not according to the previous OTU assignment.