Skip to content

Latest commit

 

History

History
342 lines (251 loc) · 12.3 KB

README.rst

File metadata and controls

342 lines (251 loc) · 12.3 KB
Raw

Command-line utilities to assist in building and publishing Galaxy tools.

image

image

image

image

Quick Start

This quick start demonstrates using planemo commands to help develop Galaxy tools.

Obtaining Planemo

The recommended approach for installing Planemo is to use Homebrew or linuxbrew. To install Planemo this way use the brew command as follows.

$ brew tap galaxyproject/tap
$ brew install planemo

Planemo can alternatively be installed via Conda:

$ conda config --add channels r
$ conda config --add channels bioconda
$ conda install planemo

For third option and a more traditional Python installation set up a virtualenv for Planemo (this example creates a new one in .venv/) and then install it with pip.

$ virtualenv .venv; . .venv/bin/activate
$ pip install planemo

Planemo is also available as a virtual appliance bundled with a preconfigured Galaxy server and set up for Galaxy tool development. You can choose from open virtualization format (OVA, .ova), Docker, or Vagrant appliances.

Planemo Basics

This quick start will assume you have a directory with one or more tool XML files. If none is available, one can be quickly created for demonstrating planemo as follows mkdir mytools; cd mytools; planemo project_init --template=demo.

Planemo can check tools containing XML for common problems and best practices using the lint command (also aliased as l).

$ planemo lint

Like many planemo commands - by default this will search the current directory and use all tool files it finds. It can be explicitly passed a path to tool files or a directory of tool files.

$ planemo l randomlines.xml

The lint command takes in additional options related to reporting levels, exit code, etc. These options are described in the docs or (like with all commands) can be accessed by passing --help to it.

$ planemo l --help
Usage: planemo lint [OPTIONS] TOOL_PATH

Once tools are syntactically correct - it is time to test. The test command can be used to test a tool or a directory of tools.

$ planemo test --galaxy_root=../galaxy randomlines.xml

If no --galaxy_root is defined, planemo will check for a default in ~/.planemo.yml and finally search the tool's parent directories for a Galaxy root directory. Planemo can also download and configure a disposable Galaxy instance for testing. Pass --install_galaxy instead of --galaxy_root.

$ planemo t --install_galaxy

Planemo will create a HTML output report in the current directory named tool_test_output.html (override with --test_output). See an example of such a report for Tophat.

Once tools have been linted and tested - the tools can be viewed in a Galaxy interface using the serve (s) command.

$ planemo serve

Like test, serve requires a Galaxy root and one can be explicitly specified with --galaxy_root or installed dynamically with --install_galaxy.

Tool Shed

Planemo can help you publish tools to the Galaxy Tool Shed. Check out Publishing to the Tool Shed.

Experimental Features

Planemo can also be used to explore some more experimental features related to Galaxy tooling - including support for Conda, Travis CI, Docker, CWL, and Homebrew.

Conda

Conda is package manager that among many other things can be used to manage Python packages. Please read a few notes regarding the setup:

  • Planemo cannot be installed via Conda and then run Galaxy inside the same Conda environment currently.
  • galaxy-lib is a dependency of Planemo and having this on the Python path prevents correct operation of Galaxy.

  • Conda dependency resolution can be used for Galaxy tools,

    This allows Galaxy to map conda recipes to requirement tags on tools

    Just install planemo normally via pip into a virtual environment or via brew and use the appropriate commands and options:

    $ planemo conda_init
$ planemo conda_install .
$ planemo test --galaxy_branch release_16.01 --conda_dependency_resolution .
$ planemo serve --galaxy_branch release_16.01 --conda_dependency_resolution .

    The test and serve commands above require the target Galaxy to be 16.01 or higher.

  • Galaxy can also simply run inside an existing conda environment and rather than using dependency resolution Conda packages can just be picked up from the PATH.

    Note: This isn't a well supported approach yet, and when possible Planemo and Galaxy should not be installed inside of Conda and the conda dependency resolution (as described above) should be used to allow tools to find Conda packages.

    To run Galaxy within a Conda environment, Planemo must be installed outside the Conda environment - via pip into a virtual environment or via homebrew. In the former case, don't place the virtualenv's bin directory on your PATH - it will conflict with Conda. Just reference Planemo directly /path/to/planemo_venv/bin/planemo test.

    To run Galaxy from within the environment you will need to install Galaxy dependencies into the conda environment. Target the development branch of Galaxy for this since it has "unpinned" dependencies that are easier to fullfill for conda.

    (conda-env-test) $ conda install numpy bx-python pysam # install the hard ones using conda
(conda-env-test) $ cd $GALAXY_ROOT
(conda-env-test) $ ./scripts/common_startup.sh --skip-venv --dev-wheels # install remaining ones using pip via Galaxy
(conda-env-test) $ cd /path/to/my/tools
(conda-env-test) $ /path/to/planemo_venv/bin/planemo test --skip_venv  .
(conda-env-test) $ /path/to/planemo_venv/bin/planemo serve --skip_venv  .

    A small test script in the Planemo source tree demonstrates this.

TravisCI

When tools are ready to be published to GitHub, it may be valuable to setup contineous integration to test changes committed and new pull requests. Travis CI is a service providing free testing and deep integration with GitHub.

The travis_init command <http://planemo.readthedocs.org/en/latest/commands.html#travis_init- command>__ will bootstrap a project with files to ease contineous integration testing of tools using a Planemo driven approach inspired by this great blog post by Peter Cock.

$ planemo travis_init .
$ # setup Ubuntu 12.04 w/tool dependencies
$ vim .travis/setup_custom_dependencies.bash
$ git add .travis.yml .travis
$ git commit -m "Add Travis CI testing infrastructure for tools."
$ git push # and register repository @ http://travis-ci.org/

In this example the file .travis/setup_custom_dependencies.bash should install the dependencies required to test your files on to the Travis user's PATH.

This testing approach may only make sense for smaller repositories with a handful of tools. For larger repositories, such as tools-devteam or tools-iuc simply linting tool and tool shed artifacts may be more feasible. Check out the .travis.yml file used by the IUC as example.

Docker

Galaxy has experimental support for running jobs in Docker containers. Planemo contains tools to assist in development of Docker images for Galaxy tools.

A shell can be launched to explore the Docker enviornment referenced by tools that are annotated with publically registered Docker images.

$ (planemo docker_shell bowtie2.xml)

For Docker containers still in development - a Dockerfile can be associated with a tool by sticking it in the tool's directory. Planemo can then build and tag a Docker image for this tool and launch a shell into it using the following commands.

$ planemo docker_build bowtie2.xml # asssumes Dockerfile in same dir
$ (planemo docker_shell --from_tag bowtie2.xml)

For more details see the documentation for the docker_build and docker_shell commands.

Common Workflow Language

Planemo includes highly experimental support for running a subset of valid Common Workflow Language (CWL) tools using a fork of Galaxy enhanced to run CWL tools.

$ planemo project_init --template cwl_draft3_spec
$ planemo serve --cwl cwl_draft3_spec/cat1-tool.cwl

Brew

The Galaxy development team was exploring (the effort has since shifted towards Conda) different options for integrating Homebrew and linuxbrew with Galaxy. One angle is resolving the tool requirements using brew. An experimental approach for versioning of brew recipes will be used. See full discussion on the homebrew-science issues page here. Information on the implementation can be found https://github.com/jmchilton/platform-brew until a more permanent project home is setup.

$ planemo brew_init # install linuxbrew (only need if not already installed)
$ planemo brew # install dependencies for all tools in directory.
$ planemo brew bowtie2.xml # install dependencies for one tool
$ which bowtie2
bowtie2 not found
$ . <(planemo brew_env --shell bowtie2.xml) # shell w/brew deps resolved
(bowtie2)$ which bowtie2
/home/john/.linuxbrew/Cellar/bowtie2/2.1.0/bin/bowtie2
(bowtie2)$ exit
$ . <(planemo brew_env bowtie2.xml) # or just source deps in cur env
$ which bowtie2
/home/john/.linuxbrew/Cellar/bowtie2/2.1.0/bin/bowtie2

For more information see the documentation for the brew and brew_env commands.