If you are looking to classify genomes according to the methodology used by the GTDB, we recommend using our companion tool GTDB-Tk instead of PhyloRank. PhyloRank is intended to aid the manual taxonomic curation of trees inferred from genomes spanning the bacterial or archaeal domain.
PhyloRank provides functionality for calculating the relative evolutionary divergence (RED) of taxa in a tree and for finding the best placement of taxonomic labels in a tree. Other functionality is in development and is currently unsupported.
The simplest way to install this package is through Bioconda:
conda install -c bioconda phylorank
Alternatively, it can be installed through PyPI:
pip install phylorank
This package makes use of the numpy, matplotlib, jinja2 (>=2.7.3), and dendropy (>=4.1.0) Python libraries. These must be install seperately. PhyloRank also uses mpld3 (>=0.2) which has been explicitly added to this package. It also required biolib (>=0.1.0), but this will be install with PhyloRank if you are doing the installation through pip.
PhyloRank requires Python 3 starting with v0.1.0
PhyloRank can calculate the relative evolutionary divergence (RED) of taxa in a tree in order to identify taxa at a given taxonomic rank that have conspicuous placements in the tree. This information can be used to refine the placement of taxa in the tree with the aim of taxa at the same rank having similar RED. RED values can be calculated using:
>phylorank outliers <input_tree> <taxonomy_file> <output_dir>
where
PhyloRank can decorate a tree with taxonomic information. This is accomplished by determining the placement for each taxa that results in the highest F-measure (see McDonald et al., 2012). If the provided taxonomy is incongruent with the topology of the tree, only the most cohesive lineage of a polyphyletic group will be labelled and it is possible that taxa at the same rank may be nested within each other (e.g., a lineage may contain two genus labels). The resulting decorated tree can be manually curated to obtain a taxonomy that is consistent with the tree topology. Alternatively, the resulting F-measure scores can be examined to assess the congruency between the taxonomy and tree topology. A tree can be decorated using:
>phylorank decorate <input_tree> <taxonomy_file> <output_tree> --skip_rd_refine
where
The taxonomy file is a simple tab-separated values file with two columns indicating the genome ID and Greengenes-style taxonomy string, e.g.:
>GCF_001687105.1 d__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Rhodobacterales;f__Rhodobacteraceae;g__Yangia;s__
Example data is provided in the example_data
directory. This tree can be decorated and the RED of each taxon establish as follows:
>phylorank decorate --skip_rd_refine ar122_r89.tree ar122_taxonomy_r89.tsv ar122_r89.decorated.tree
>phylorank outliers ar122_r89.decorated.tree ar122_taxonomy_r89.tsv output_dir
If you find this package useful, please cite this git repository (https://github.com/dparks1134/PhyloRank)
Copyright © 2015 Donovan Parks. See LICENSE for further details.