284 ◾ Bioinformatics
qiime metadata tabulate \
--m-input-file dada2/stats_yoga_dada2.qza \
--o-visualization dada2/stats_yoga_dada2.qzv
qiime tools view dada2/stats_yoga_dada2.qzv
The DADA2 stats summary table is a TSV. It shows the ID, number of input reads, number
of reads filtered, percentage of reads passed the filter, number of denoised reads, number of
merged reads, percentage of reads merged, number of non-chimeric reads, and percentage
of non-chimeric reads for each sample in the study.
7.3.4.2.2.2 Denoising with Deblur
Deblur plugin denoises single-end reads and merged paired-end read. Paired-end sequences
must be merged using “vsearch join-pairs”, as shown above, before being denoised by
deblur.
mkdir deblur
qiime vsearch join-pairs \
--i-demultiplexed-seqs inputs/demux-yoga.qza \
--p-allowmergestagger \
--o-joined-sequences deblur/demux-yoga-merged.qza
It is recommended to run quality assessment and quality control by filtering low-quality
reads using “quality-filter q-score” before denoising with deblur.
qiime quality-filter q-score \
--i-demux inputs/demux-yoga-merged.qza \
--o-filtered-sequences deblur/demux-yoga-merged-filtered.qza \
--o-filter-stats deblur/demux-yoga-merged-filtered-stat.qza
Denoising with deblur is carried out with “q2-deblur” plugin, which has two methods:
“denoise-16S” for denoising 16S rRNA gene sequences and “denoise-other” for denoising
other types of targeted gene sequences. When you use “denoise-16S”, deblur will perform
an initial positive filtering step by discarding any reads that do not have a minimum 60%
identity similarity to database sequences from the 85% GreenGenes.
You can use “deblur denoise-16S” method, which has “--p-trim-length” parameter that
can be set to truncate the reads at specific position for quality filtering or you can set that
parameter to -1 to disable trimming. The following script will perform deblur denoising:
mkdir deblur
qiime deblur denoise-16S \
--i-demultiplexed-seqs deblur/demux-yoga-merged.qza \
--p-trim-length -1 \
--p-jobs-to-start 4 \
--p-sample-stats \
--o-representative-sequences deblur/rep-seqs_yoga_deblur.qza \