taxtastic 0.9.3

Tools for taxonomic naming and annotation

Taxtastic is a python package used to build and maintain reference packages– i.e. collections of reference trees, reference alignments, profiles, and associated taxonomic information.

We love it, but what is it?

• quickstart

• full documentation

A script named taxit provides a command line interface:

% taxit  --help
usage: taxit [-h] [-V] [-v] [-q]
             {help,add_nodes,add_to_taxtable,check,composition,create,extract_nodes,findcompany,get_lineage,info,lonelynodes,new_database,refpkg_intersection,reroot,rollback,rollforward,rp,strip,taxids,taxtable,update,update_taxids}
             ...

Creation, validation, and modification of reference packages for use with
`pplacer` and related software.

positional arguments:
  {help,add_nodes,add_to_taxtable,check,composition,create,extract_nodes,findcompany,get_lineage,info,lonelynodes,new_database,refpkg_intersection,reroot,rollback,rollforward,rp,strip,taxids,taxtable,update,update_taxids}
    help                Detailed help for actions using `help <action>`
    add_nodes           Add nodes and names to a database
    add_to_taxtable     Add nodes to an existing taxtable csv
    check               Validate a reference package
    composition         Show taxonomic composition of a reference package
    create              Create a reference package
    extract_nodes       Extract nodes from a given source in yaml format
    findcompany         Find company for lonely nodes
    get_lineage         Calculate the taxonomic lineage of a taxid
    info                Show information about reference packages.
    lonelynodes         Extracts tax ids of all lonely nodes in a taxtable
    new_database        Download NCBI taxonomy and create a database
    refpkg_intersection
                        Find the intersection of a taxtable and a refpkg's
                        taxonomy.
    reroot              Taxonomically reroots a reference package
    rollback            Undo an operation performed on a refpkg
    rollforward         Restore a change to a refpkg immediately after being
                        reverted
    rp                  Resolve path; get the path to a file in the reference
                        package
    strip               Remove rollback and rollforward information from a
                        refpkg
    taxids              Convert a list of taxonomic names into a recursive
                        list of species
    taxtable            Create a tabular representation of taxonomic lineages
    update              Add or modify files or metadata in a refpkg
    update_taxids       Update obsolete tax_ids

optional arguments:
  -h, --help            show this help message and exit
  -V, --version         Print the version number and exit
  -v, --verbose         Increase verbosity of screen output (eg, -v is
                        verbose, -vv more so)
  -q, --quiet           Suppress output

Installation

taxtastic requires Python 3.5+. Python 2 is no longer supported. The simplest method of installing is using pip:

pip install taxtastic

We strongly recommend installation into a virtualenv. On a clean Ubuntu 16.04+ system, complete instructions for installing the taxtastic package and the taxit command line entry point in a virtualenv are below.

Create a virtualenv and install taxtastic:

python3 -m venv taxtastic-env
source taxtastic-env/bin/activate
pip install -U pip
pip install taxtastic

If you prefer to install from the git repository:

git clone https://github.com/fhcrc/taxtastic.git
cd taxtastic
python -m venv taxtastic-env
source taxtastic-env/bin/activate
pip install .

If you want to live dangerously and install the package to the system despite our pleas not to do so:

sudo apt-get install python-pip
sudo pip install taxtastic

If you are not familiar with python virtual environments, the following post is helpful: https://realpython.com/blog/python/python-virtual-environments-a-primer/

Finally, taxit can be run from a docker image hosted from Docker Hub (https://hub.docker.com/r/nghoffman/taxtastic/), for example:

docker run --rm -it -v $(pwd):$(pwd) -w $(pwd) nghoffman/taxtastic:latest taxit -v new_database

sqlite3

Taxtastic uses recursive common table expressions to query the taxonomy database, which requires that the Python sqlite3 module is built against sqlite3 library version of 3.8.3 or higher (http://www.sqlite.org/releaselog/3_8_3.html). You can check the version like this:

python -c 'import sqlite3; print sqlite3.sqlite_version'

python setup.py will exit with an error if the sqlite3 library dependency is not met. On older systems (and for python2 only), it is possible to replace the builtin sqlite3 module by installing pysqlite2 with updated sqlite3 libraries before installing the package with pip using a provided script in the taxtastic git repository (assuming an active virtualenv):

dev/install_pysqlite.sh

Or, to avoid cloning the repo:

curl https://raw.githubusercontent.com/fhcrc/taxtastic/master/dev/install_pysqlite.sh | bash

After the script completes, confirm that pysqlite2 was installed:

python -c 'from pysqlite2 import dbapi2; print dbapi2.sqlite_version'

At this point, taxtastic may be installed as described above.

Note that pysqlite2 is available for python2 only, so there really is no good option for using python3 on older systems like Ubuntu 14.04 and earlier, unless you want to compile a version of the python3 interpreter linked against updated sqlite3 libraries. If you must use an older system, stick with python2, or use the Docker image.

A note on databases

This project supports both sqlite3 and postgresql as database backends. For most applications, we recommend sqlite3: some operations (particularly initial database creation) are much faster using sqlite3 due to the details of how postgresql enforces database constraints (we may try to optimize this in the future - in theory, postgresql can be made to be at least as fast). If you do want to use postgresql, note that some of the queries consume a lot of memory, and the default configuration tends to be memory constrained (and this really slows things down). On a reasonably new mac laptop, we found that the optimizations suggested here (http://big-elephants.com/2012-12/tuning-postgres-on-macos/) do the trick.