Dada2 parameters - low percentage of input non-chimeric

Hello everyone
I am new here and I’ve read lots of posts about dada2 parameters but I couldn’t find any answer for me.

I used 515F–806R primers to get amplicons. Samples were extracted from chicken feces swabs. Reads are paired-end 2x300bp. Running QIIME2 2020.6 in a conda environment.

Based on the quality score plot above, I used the following code in the Dada2 step:

qiime dada2 denoise-paired /
–i-demultiplexed-seqs new-analysis-kit_comparison/pe-kit-demux_kit1.qza /
–p-trim-left-f 0 /
–p-trunc-len-f 260 /
–p-trunc-len-r 274
–o-representative-sequences rep-seqs-kit1.qza /
–o-table table-kit1-dada2.qza /
–o-denoising-stats stats-dada2-kit1.qza /
–p-n-threads 0 /

As you can see on stats data file, I think I had a good filter and merge results, but I got a low percentage of input non-chimeric, which I think means that I lost 90% of my data.
Any clue on what is happening?

Thanks in advance

demux-kit1.qzv (316.1 KB)

stats-dada2-kit1.qzv (1.2 MB)

1 Like

Hi @natalia_gaeta ,
Wow, you are losing a lot of reads, definitely an abnormally high number! You should check out these related forum topics for more discussion on how to troubleshoot and adjust… one thing to make sure is that you are removing your primers prior to dada2 (either with q2-cutadapt or with the “trim-left” parameters in dada2:

Good luck!

1 Like

Thanks Nicholas! I will check them and give you any updates.

1 Like