Contents

Downstream alignment and phylogenetics with QIIME 2

QIIME 2 contains tools for sequence alignment, alignment filtering, and phylogenetic reconstruction in the q2-alignment and q2-phylogeny plugins (which are not installed by default with genome-sampler). This document illustrates how these steps can be applied after working through genome-sampler usage tutorial.

QIIME 2 often wraps other widely used tools in QIIME 2 plugins rather than implement them directly. For example, under-the-hood, genome-sampler’s sample-diversity action is using vsearch [1]. In this document we’ll build an alignment with MAFFT [2], apply a pre-computed alignment position mask, and then build a phylogenetic tree with IQTree 2 [3].

Warning

QIIME 2 is not designed for high-accuracy phylogenetic analysis. If you use the workflow illustrated here, we recommend that you consider the resulting trees to be experimental. In other words, this workflow can give you a quick idea of what is happening with your data, but you may want to use other tools for your final trees.

Note

This tutorial assumes that you have already run genome-sampler usage tutorial. If you haven’t, you’ll be missing files that are needed for this tutorial.

Obtain reference sequence and alignment mask

curl -sL https://raw.githubusercontent.com/caporaso-lab/genome-sampler/r2020.8/snakemake/tutorial-data/alignment-mask.qza --output tutorial-data/alignment-mask.qza
curl -sL https://raw.githubusercontent.com/caporaso-lab/genome-sampler/r2020.8/snakemake/tutorial-data/sarscov2-reference-genome.qza --output tutorial-data/sarscov2-reference-genome.qza

Align sequences and build a tree

First, we’ll add the SARS-CoV-2 reference sequence to the sequence collection obtained in the tutorial. Notice that we’re working with the .qza file that was created in that tutorial, not the .fasta file that we exported.

qiime feature-table merge-seqs \
  --i-data sequences.qza \
  --i-data sarscov2-reference-genome.qza \
  --o-merged-data sequences-w-ref.qza

Next, we’ll perform sequence alignment using MAFFT.

qiime alignment mafft \
  --i-sequences sequences-w-ref.qza \
  --o-alignment aligned-sequences-w-ref.qza

After aligning the sequences, a “mask” can be applied to filter positions from the alignment that are likely to be uninformative. At present, we’re experimenting with a pre-computed alignment mask.

qiime genome-sampler mask \
  --i-alignment aligned-sequences-w-ref.qza \
  --i-mask alignment-mask.qza \
  --o-masked-alignment masked-aligned-sequences-w-ref.qza

Finally, we build a tree from the resulting alignment. This will generate an unrooted phylogenetic tree.

qiime phylogeny iqtree \
  --i-alignment masked-aligned-sequences-w-ref.qza \
  --o-tree unrooted-tree.qza

This .qza file can be viewed directly with iTOL to get a quick look. We’ll also soon be adding support for viewing this with Empress.

All of the .qza files that were generated in this example can be exported using qiime tools export. Exporting of sequence or alignment files will provide you with fasta files by default, and exporting of the phylogenetic tree will provide you with a newick file by default. See the QIIME 2 exporting documentation for more details.

genome-sampler usage tutorial Other topics