QIIME 2, mothur, and BioConductor

QIIME 2

rbiom ➡ QIIME 2

In R, export QIIME 2-compatible files.

library(rbiom)

# where to save the files
project_dir <- tempdir()

write_qiime2(biom = hmp50, dir = project_dir, prefix = 'hmp50_')

This command creates files named 'hmp50_counts.tsv', 'hmp50_metadata.tsv', 'hmp50_taxonomy.tsv', 'hmp50_tree.nwk', and 'hmp50_seqs.fna'.

On the command line, convert files to QIIME 2 format (.qza).

# Convert classic BIOM table to HDF5
biom convert -i hmp50_counts.tsv -o hmp50_counts.hdf5 --to-hdf5

# Import counts
qiime tools import \
  --input-path hmp50_counts.hdf5 \
  --type 'FeatureTable[Frequency]' \
  --input-format BIOMV210Format \
  --output-path hmp50-counts.qza

# Import taxonomy
qiime tools import \
 --input-path hmp50_taxonomy.tsv \
 --type FeatureData[Taxonomy] \
 --output-path hmp50-taxonomy.qza

# Import phylogenetic tree
qiime tools import \
  --input-path hmp50_tree.nwk \
  --type 'Phylogeny[Rooted]' \
  --output-path hmp50-tree.qza

Examples: running unifrac and browsing metadata

qiime diversity-lib weighted-unifrac \
  --i-table hmp50-counts.qza \
  --i-phylogeny hmp50-tree.qza \
  --o-distance-matrix weighted-unifrac-dm.qza
  
qiime metadata tabulate \
  --m-input-file hmp50_metadata.tsv \
  --o-visualization hmp50-metadata-browser.qzv
# View .qzv files with https://view.qiime2.org

QIIME 2 ➡ rbiom

On the command line, export data files from QIIME 2.

# Export a feature table, taxonomy, and tree
qiime tools export --input-path hmp50-counts.qza   --output-path .
qiime tools export --input-path hmp50-taxonomy.qza --output-path .
qiime tools export --input-path hmp50-tree.qza     --output-path .

These commands create files named 'feature-table.biom', 'taxonomy.tsv', and 'tree.nwk'.

In R, import the data files into rbiom.

# project_dir = directory with relevant files

withr::with_dir(project_dir, {

  biom <- as_rbiom(
    biom     = 'feature-table.biom', 
    metadata = 'hmp50_metadata.tsv', 
    taxonomy = 'taxonomy.tsv', 
    tree     = 'tree.nwk' )
  
})

mothur

rbiom ➡ mothur

In R, export mothur-compatible files.

library(rbiom)

# where to save the files
project_dir <- tempdir()

write_mothur(biom = hmp50, dir = project_dir, prefix = 'hmp50_')

This command creates files named 'hmp50_counts.tsv', 'hmp50_metadata.tsv', 'hmp50_taxonomy.tsv', 'hmp50_tree.nwk', and 'hmp50_seqs.fna'.

At the mothur command prompt, import the BIOM data.

mothur > make.shared(count=hmp50_counts.tsv, label=asv)
mothur > unifrac.unweighted(tree=hmp50_tree.nwk, count=hmp50_counts.tsv)

The make.shared() command creates files named 'hmp50_counts.asv.list' and 'hmp50_counts.asv.shared'.

mothur ➡ rbiom

At the mothur command prompt, export a biom file.

mothur > make.biom( \
  shared=hmp50_counts.asv.shared, \
  constaxonomy=hmp50_taxonomy.tsv, \
  metadata=hmp50_metadata.tsv, \
  output=simple )

The make.biom() command creates a file named 'hmp50_counts.asv.asv.biom'.

In R, import the biom file with rbiom.

# project_dir = directory with relevant files

withr::with_dir(project_dir, {

  biom <- as_rbiom('hmp50_counts.asv.asv.biom')

  # Optionally add a tree and/or sequences
  biom$tree <- 'hmp50_tree.nwk'
  biom$seqs <- 'hmp50_seqs.fna'
  
})

BioConductor R Packages

rbiom ➡ phyloseq

# An rbiom object
biom <- rbiom::hmp50

# A phyloseq object
physeq <- rbiom::convert_to_phyloseq(biom)

phyloseq ➡ rbiom

# A phyloseq object
data(enterotype, package = 'phyloseq')
physeq <- enterotype

# An rbiom object
biom <- rbiom::as_rbiom(physeq)

rbiom ➡ SummarizedExperiment

# An rbiom object
biom <- rbiom::hmp50

# A SummarizedExperiment object
se <- rbiom::convert_to_SE(biom)

SummarizedExperiment ➡ rbiom

# `se` is a SummarizedExperiment object

# Convert to rbiom object
biom <- rbiom::as_rbiom(se)

rbiom ➡ TreeSummarizedExperiment

# An rbiom object
biom <- rbiom::hmp50

# A TreeSummarizedExperiment object
tse <- rbiom::convert_to_TSE(biom)

TreeSummarizedExperiment ➡ rbiom

# `tse` is a TreeSummarizedExperiment object

# Convert to rbiom object
biom <- rbiom::as_rbiom(tse)