Pomoxis - bioinformatics tools for nanopore research

© 2018 Oxford Nanopore Technologies Ltd.

Pomoxis comprises a set of basic bioinformatic tools tailored to nanopore sequencing. Notably tools are included for generating and analysing draft assemblies. Many of these tools are used by the research data analysis group at Oxford Nanopore Technologies.

See Pomoxis Programs for common simple tasks.

Features

  • Wraps third party tools with known good default parameters and methods of use.

  • Creates an isolated environment with all third-party tools.

  • Can be installed with conda.

  • Streamlines common short analysis chains.

  • Integrates into katuali for performing more complex analysis pipelines.

  • Open source (Mozilla Public License 2.0).

Research Release

Research releases are provided as technology demonstrators to provide early access to features or stimulate Community development of tools. Support for this software will be minimal and is only provided directly by the developers. Feature requests, improvements, and discussions are welcome and can be implemented by forking and pull requests. However much as we would like to rectify every issue and piece of feedback users may have, the developers may have limited resource for support of this software. Research releases may be unstable and subject to rapid iteration by Oxford Nanopore Technologies.

Compatibility

Pomoxis is developed on Ubuntu 16.04, other recent linuxes should be equally compatible (see build notes below). Pomoxis is known to work on at least some MacOS High Sierra configurations, though some components, notably scrappy, are known to not work on some MacOS configurations (combinations of OS and xcode versions).

Installation

Much of pomoxis’s functionality is dependent on thirty party tools. These can be provided by the user, or can be installed with the help of the provided Makefile

Installation with conda

Pomoxis is available on bioconda and so can be most easily installed with:

conda install pomoxis

Installation with pip

Pomoxis can be installed using the python package manager, pip:

pip install git+https://github.com/rrwick/Porechop
pip install pomoxis

We recommend using medaka within a virtual environment, viz.:

virtualenv pomoxis --python=python3 --prompt "(pomoxis) "
. pomoxis/bin/activate
pip install git+https://github.com/rrwick/Porechop
pip install pomoxis

Note

Using this method requires the user to provide several binaries:

minimap2, miniasm, samtools, racon, samtools, bcftools, and seqkit,

and place these within the PATH.

Installation from source

With this method pomoxis will install itself into a an isolated virtual environment. The installation will fetch, compile, and install all direct dependencies into the environment. Use this method if you do not wish to use conda, but will not be providing the third-party binaries.

Note

Before installing pomoxis is may be required to install some prerequisite packages, best installed by a package manager. On Ubuntu these are:

gcc-4.9 g++-4.9 zlib1g-dev libncurses5-dev python3-all-dev libhdf5-dev libatlas-base-dev libopenblas-base libopenblas-dev libbz2-dev liblzma-dev libffi-dev make python-virtualenv cmake wget bzip2

To setup the environment run:

git clone --recursive https://github.com/nanoporetech/pomoxis
cd pomoxis
make install
. ./venv/bin/activate

The installation of porechop (https://github.com/rrwick/Porechop) requires a newer compiler than is a available on some systems. It may therefore be necessary to install a newer compiler and set environment variables before the make install step:

# For porechop to be compiled on older systems set these, e.g.:
export CXX="g++-4.9" CC="gcc-4.9"

Note also that racon requires at least gcc>=4.8.5 to [compile smoothly](https://github.com/isovic/racon/issues/57).

Installation without compiling third-party binaries

Running the above within a pre-exisiting (virtual) environnment may well fail; advanced may wish to simply run

python setup.py install

in the standard manner after compiling the third party programs listed below and ensuring they are present on the PATH. The setup.py script can copy the binaries into the python interpreter path if they are placed within a directory named bincache alongside setup.py. To make use of this facility run:

pip install -r requirements.txt
POMO_BINARIES=1 python setup.py install

Third party binaries

The distribution bundles some common bioinformatics tools:

  • miniasm

  • minimap2

  • racon

  • samtools

  • bcftools

  • seqkit

  • porechop

Indices and tables