QIIME2 DADA2 Quick Reference

There are detailed tutorials for how to run DADA2 with QIIME 2 on the QIIME 2 website. This page is simply a quick reference for convenience. Running the DADA2 pipeline is more straight-forward than deblur, but is slower. It is also more flexible with the truncation lengths, but note that these are specified for forward and reverse reads separately. The important thing to consider here is what lengths to use so that the forward and reverse reads can be stitched together accurately. You may also want to lower the maximum number of expected errors allowed depending on your data quality.

Here is a quick reference for running DADA2 in QIIME2 (note that you may want to decrease the number of threads).

qiime dada2 denoise-paired --i-demultiplexed-seqs reads_qza/reads_trimmed.qza \
                           --p-trunc-len-f 270 \
                           --p-trunc-len-r 210 \
                           --p-max-ee-f 2 \
                           --p-max-ee-r 3 \
                           --p-n-threads 30 \
                           --output-dir dada2_output

You should then take a look at the read count table to see how many reads were retained at each step of the DADA2 pipeline:

qiime tools export --input-path dada2_output/denoising_stats.qza --output-path dada2_output