Feature count after DADA2/Deblur ~ most of initial features seems low to me(paired end reads)

Based on the demux.qzvqualityplots2 qualityplots2 s, how can I determine the best specific position using the parameters (i.e. --p-trunc-len-f and --p-trunc-len-r) in the command qiime dada2 denoise-paired. Actually, the quality plots (prior to quality filtering) suggest a reasonable choice is in a kind of sequence position range, not a specific position. This is a little subjective assessment and tricky.

Below is my quality plots for forward, reverse and joined single reads. Could you please give me some suggestions on the selection of the length to make sure that most of the reads can pass the step?



Hi @tengfey,
A couple of things to clarify first.
The title here suggests that you have ran into issues of low features making it through DADA2 and Deblur? If this is the case could you also share the stats.qzv produced by those steps.
You’ve posted 2 identical images of your forward & reverse quality scores, was one of them meant to be something else?
The question you ask regarding truncating/trimming parameters has been discussed and answered extensively in the forum, I recommend browsing around to some previous posts for good tips. For paired-end reads there is specifically an important consideration of retaining a min 20bp overlap region after truncating to allow for proper merging in DADA2. What are the primer sets used and the expected amplicon length here?

1 Like

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