ITS denoising truncation

I understand working with ITS reads can be tricky, but I was looking for a second opinion on truncation. I have tried to denoise at many truncation lengths but I seem to only get 30-40% input in my denoising summary.

denoising-stats5.qzv (1.2 MB)

this is my most recent code that gave me the best results:

qiime cutadapt trim-paired \
--i-demultiplexed-sequences** /scratch/rylee650/demux-paired-end.qza \
--p-front-f ACACTCTTTCCCTACACGACGCTCTTCCGATCTCTTGGTCATTTAGAGGAAGTAA \
--p-front-f ACACTCTTTCCCTACACGACGCTCTTCCGATCTTCTTGGTCATTTAGAGGAAGTAA \
--p-front-f ACACTCTTTCCCTACACGACGCTCTTCCGATCTACCTTGGTCATTTAGAGGAAGTAA \
--p-front-f ACACTCTTTCCCTACACGACGCTCTTCCGATCTCAACTTGGTCATTTAGAGGAAGTAA \
--p-front-r GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGCTGCGTTCTTCATCGATGC \
--p-front-r GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTTGCTGCGTTCTTCATCGATGC \
--p-front-r GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTACGCTGCGTTCTTCATCGATGC \
--p-front-r GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTCATGCTGCGTTCTTCATCGATGC \
--p-adapter-f GCATCGATGAAGAACGCAGC \
--p-adapter-r TTACTTCCTCTAAATGACCAAG \
--p-discard-untrimmed \
--o-trimmed-sequences /scratch/rylee650/demux-trimmed.qza

qiime demux summarize \
--i-data /scratch/rylee650/demux-trimmed.qza \
--o-visualization /scratch/rylee650/demux-trimmed.qzv

qiime dada2 denoise-paired \
--i-demultiplexed-seqs /scratch/rylee650/demux-trimmed.qza \
--p-trim-left-f 0 \
--p-trim-left-r 0 \
--p-trunc-len-f 278 \
--p-trunc-len-r 225 \
--o-table /scratch/rylee650/table5.qza \
--o-representative-sequences /scratch/rylee650/rep-seqs5.qza \
--o-denoising-stats /scratch/rylee650/denoising-stats5.qza

qiime metadata tabulate \
--m-input-file /scratch/rylee650/denoising-stats5.qza \
--o-visualization /scratch/rylee650/denoising-stats5.qzv

my question then, is at what point am I truncating too much, and will it be possible to reach above 50% passing the input filter?

I have attached my demux trimmed and original demux file, feel free to take a look and see if everything looks proper in those files.

demux-trimmed.qzv (319.9 KB)

demux.qzv (313.6 KB)

TIA

Hi @rylee650,

One additional option you can try is to provide the reverse compliment of the primers and adapters for the opposite reads. Analogous to this thread. This will help trim sequence from the 3' end of the given read that might result in "read-through" into the primer and adapter sequence.

1 Like

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