288    ◾    Bioinformatics

After you study the feature table visualization on the Internet browser, you may decide

to remove some samples or features from the feature table because of outliers or low abun-

dance. The “q2-feature-table” plugin has the “filter-samples” and “filter-features” methods

for these purposes. For example, you can remove samples based on their minimum total

frequency. The following script removes the samples with a total frequency less than 1000

from feature table created with DADA2 denoising and then it creates a visualization and

views it on the Internet browser:

qiime feature-table filter-samples \

--i-table dada2/table_yoga_dada2.qza \

--p-min-frequency 1000 \

--o-filtered-table dada2/table_sample_freq_filtered_yoga_dada2.

qza

qiime feature-table summarize \

--i-table dada2/table_sample_freq_filtered_yoga_dada2.qza \

--m-sample-metadata-file data/sample-metadata.tsv \

--o-visualization dada2/table_sample_freq_filtered_yoga_dada2.qzv

qiime tools view dada2/table_sample_freq_filtered_yoga_dada2.qzv

Try to study the feature table report after the above filtering.

The “filter-features” method is used to remove low abundance features from a feature

table. The following script removes features with a total abundance (across all samples) of

less than 20:

qiime feature-table filter-features \

--i-table dada2/table_sample_freq_filtered_yoga_dada2.qza \

--p-min-frequency 20 \

--o-filtered-table dada2/table_feat_sample_freq_filtered_yoga_

dada2.qza

FIGURE 7.14  The tabular report of the amplicon sequence variants (ASVs).