How to deal with sequencing with relatively poor quality?

Hello, QIIME team,

I newly received sequencing results for my 150 samples using 16S rRNA sequencing. From the interactive Quality Plot report (qiime2), I could see the quality dropped below 20 after ~210 bp for the forward reads, while for the reverse reads, the quality score dropped below 20 after ~160 bp. I am using the primer pair 341F and 805R. I think this primer pair covers 464 bp length by using 805-341.

I used the following command for denosing and merging forward and reverse reads

“qiime dada2 denoise-paired --i-demultiplexed-seqs 16S_paired-end-demux.qza --p-trunc-len-f 210 --p-trunc-len-r 160 --p-trim-left-f 17 --p-trim-left-r 21 --p-n-threads 16 --o-table 16S_table.qza --o-representative-sequences 16S_seqs.qza”

The sequences before denosing was ~1600,000 bp, but was only 129 after denoising.

So I guess, there were only a few sequences merged and passed the denoising procedure.

Now, in my case, should I only use the forward reads to continue the analysis or there is still possibility to use both forward and reverse reads?

And, I want to know should I must remove the template specific primers from 16S and ITS sequences before further processing? And why?


Hey @hongwei2017!

According to your parameters, you would have:

210 - 17 = 206 bases going forward and 160 - 21 = 139 in the reverse direction. This gives you 345 bases optimistically (and DADA2 expects 20 bases of overlap to merge). So I don’t think you are going to be able to merge if your primer-pair is generally ~464 bp. You are roughly 120 (plus 20) bases short. There’s nothing wrong with relaxing your trunc-len a little bit to see if you can get enough overlap, but I suppose I don’t know how bad your quality gets by the time you would make up for the missing ~140 bases.

You may be better off going with forward, but it’s ultimately up to you.

DADA2 expects biological sequences, and so anything like a primer/linker/adapter will produce poor results. This may also be a factor in your low merge counts (but I think lack of overlap is more likely).

Usually you can use the trim-left to get rid of the primers, but in the case of ITS, sometimes you get the reverse primer on your forward reads, and that is just as troublesome for DADA2 as your forward primer.

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