All features filtered out even though input sequence quality is good

Hi, all! I'm currently working with 16S V3-V4 region data. I supposed to do dada2 denoise-paired, but most of samples are filtered out on merge stage, so I decided to only use forward read. Below is the quality of sequence and command for dada2.

qiime dada2 denoise-single
--i-demultiplexed-seqs filtered_sequences.qza
--p-trim-left 7
--p-trunc-len 288
--p-n-threads 15
--p-max-ee 30
--o-table denoise_outcome/table.qza
--o-representative-sequences denoise_outcome/rep-seqs.qza
--o-denoising-stats denoise_outcome/stats.qza

But after running it without any error message, I got below result:

Almost all of sample reads are filtered out even if I changed --p-max-ee value (25~30). Before running it, input sequences are quality filtered with command below (so I used the filtered-sequences.qza file for denoise-single from below job)

qiime quality-filter q-score
--i-demux demux.qza
--p-min-quality 28
--output-dir demux_flted

Any idea I can resolve it?
Thank you in advance!

Is your read length 240 (appears that way from top figure). If yes, then reduce trunc-len to below 240, maybe 200.

Hi, @kindergarten !

The end of picture was cut somehow (the full length of input is 300bp).
But your opinion works perfectly! I tried to keep the read as long as possible, so I didn't think of reduce trunc-len value that much.

Thank you :slight_smile:

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