q2-shared_asv is a QIIME 2 plugin for shared ASV analysis.
# Activate your qiime2
conda activate qiime2-2023.2
pip install git+https://github.com/biota-inc/q2-shared_asv.git
qiime dev refresh-cache
qiime --help$ qiime shared-asv compute --help
Inputs:
--i-table ARTIFACT FeatureTable[RelativeFrequency]
The feature table containing the samples for which
shared ASVs should be computed. [required]
Parameters:
--p-sample-a TEXT The first sample for which shared ASVs should be
computed. [required]
--p-sample-b TEXT The second sample for which shared ASVs should be
computed. [required]
--m-metadata-file METADATA...
(multiple The sample metadata for sample-id
arguments will
be merged) [required]
--p-percentage PROPORTION Range(0, 1, inclusive_end=True)
The threshold for filtering shared ASVs based on
relative frequency. [required]
Outputs:
--o-shared-asvs ARTIFACT FeatureTable[RelativeFrequency]
The resulting feature table containing the shared ASVs
between the two samples. [required]For 16S rRNA amplicon sequencing, use biota-inc/16S_pipeline on Github and create analysis/table-no-unassigned-feature-filtered.qza.
- Make a relative frequency table
# Compute the relative frequency of the features
qiime feature-table relative-frequency \
--i-table analysis/table-no-unassigned-feature-filtered.qza \
--o-relative-frequency-table analysis/relative_frequency.qza
# Summarize the relative frequency table
qiime feature-table summarize \
--i-table analysis/relative_frequency.qza \
--o-visualization analysis/relative_frequency.qzv- Make a table like below and name it as shared_asv.txt (tab-demilited format txt file).
Note: An empty last line is required!
| Pair A | Pair B | Pair ID |
|---|---|---|
| S1 | N1 | 1 |
| S2 | N2 | 2 |
| S3 | N3 | 3 |
| S4 | N4 | 4 |
| S5 | N5 | 5 |
- Run the command below. This step generates the shared ASV table of each pair.
# If tab delimiters are not recognized well, enclose `$line` and `$PairA(B)` with Quotation Marks (for example, `$(echo "$line" |`)
tail -n +2 metadata/shared_asv.txt | while read line; do
PairA=$(echo $line | awk -F'\t' '{print $1}' | tr -d '[:space:]')
PairB=$(echo $line | awk -F'\t' '{print $2}' | tr -d '[:space:]')
ID=$(echo $line | awk -F'\t' '{print $3}' | tr -d '[:space:]')
qiime shared-asv compute \
--i-table relative_frequency.qza \
--m-metadata-file metadata/sample-data.txt \
--p-sample-a $PairA \
--p-sample-b $PairB \
--p-percentage 0.0001 \
--o-shared-asvs analysis/shared-asvs_$ID.qza
done- Filter samples for each shared-asvs file
# For example, filtering samples by using sample-data_skin.txt extracting only skin label of sample_type (if not, create this kind of file)
for i in {1..39}; do
qiime feature-table filter-samples \
--i-table analysis/shared-asvs_${i}.qza \
--m-metadata-file metadata/sample-data_skin.txt \
--o-filtered-table analysis/shared-asvs_${i}_skin.qza
done- Merge the table files into one!
# Initialize the merged table with the first shared-asvs skin file
cp analysis/shared-asvs_1_skin.qza analysis/merged-table.qza
# Merge all shared-asvs skin files
for i in {2..39}; do
qiime feature-table merge \
--i-tables analysis/merged-table.qza \
--i-tables analysis/shared-asvs_${i}_skin.qza \
--o-merged-table analysis/temp_merged-table.qza
mv analysis/temp_merged-table.qza analysis/merged-table.qza
done- Final steps
# Summarize the merged table
qiime feature-table summarize \
--i-table analysis/merged-table.qza \
--o-visualization analysis/merged-table.qzv
# Export the merged table as a .biom file
qiime tools export \
--input-path analysis/merged-table.qza \
--output-path analysis/merged-table
# Convert the .biom file to a TSV format
biom convert \
-i analysis/merged-table/feature-table.biom \
-o analysis/merged-table/table_biom.txt \
--to-tsvq2-shared_asv is available under the BSD-3-Clause License. See the LICENSE file for more info.