After dada2 denoising most of samples have feature count zero

[Daria]

Hello,
this is my first experience with Qiime2. I try processing fungal ITS data. I have demultiplexed and trimmed data. As I understood it is Casava 1.8 paired-end demultiplexed format (my data: 1K1_S1_L001_R1_001.fastq.gz 1K1_S1_L001_R2_001.fastq.gz). I have 45 samples.
But after denoising by dada2 32 of 45 samples have Feature Count zero. Also when I checked denoising-status I saw that a half of samples haven't married sequences. I need chose sample depth for diversity analysis and if I exclude all samples with depth 0, there will left only 13 samples.
I've already tried changing the values of p-trim-left and p-trunc-len but it didn't help, the result was same.

qiime tools import
--type 'SampleData[PairedEndSequencesWithQuality]'
--input-path /home/daria/Qiime2/Bielsk
--input-format CasavaOneEightSingleLanePerSampleDirFmt
--output-path demux-paired-end.qza

qiime dada2 denoise-paired
--i-demultiplexed-seqs demux-paired-end.qza
--p-trim-left-f 7
--p-trim-left-r 7
--p-trunc-len-f 140
--p-trunc-len-r 140
--o-table table.qza
--o-representative-sequences rep-seqs.qza
--o-denoising-stats denoising-stats.qza

I suppose that I made mistake in the denoising step, because table with denoising status looks very strange, or may be I need another plugin for denoising. Please explain me what I did wrong.

Can anyone help me?
Thanks a lot!

[jwdebelius (Justine Debelius)]

Hi @Daria,

It looks like your samples are failing in read merging. I’m not sure what your expected amplicon length is, but 2 x 140 does not seem long enough to merge the reasons. So, I’d probably proceed with your forward reads only because it looks like those are coming through the denoising just fine.

Best,
Justine

[Daria]

Thank you so much. You was absolutely right, I tried with only forward reads and got good result