Biopython package definitions for the Galaxy Tool Shed (OBSOLETE)
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


Biopython package definitions for the Galaxy Tool Shed (OBSOLETE)

For an introduction to the Biopython Project, please see - and for an introduction to the Galaxy Project, please see

The Galaxy web platform is extensible allowing addition of new tools to a Galaxy instance. These tools are shared via a Galaxy Tool Shed, primarily and the Test Tool Shed

In addition to new tools, the Galaxy Tool Shed also included software package definitions for use as dependencies. This repository was for the development of the Biopython package definitions shared on the main Tool Shed and Test Tool Shed:

Note that each release of Biopython gets its own predicatably named entry on the Tool Shed, so that tools using Biopython can specify the exact version of Biopython required. This is important for reproducibility which is a Galaxy design goal.

However, since 2017 the Galaxy community has moved away from this home-grown dependency system and adopted Conda, and in particular the BioConda channel. This covers Biopython:

Instructions for Galaxy Tool Authors

If you are writing a Galaxy Tool which uses a Python script dependent on Biopython, in order for the Galaxy Tool Shed to automatically install Biopython you must include a tool_dependencies.xml file declaring this as a dependency, e.g.:

<?xml version="1.0"?>
<repositories description="Requires Biopython as a dependency.">
  <repository name="package_biopython_1_65" owner="biopython" />

Also, in your tool XML file(s) you must include:

  <requirement type="package" version="1.65">biopython</requirement>

This will work on the main Tool Shed or the Test Tool Shed where it will refer to these packages:

Here is an example declaring a dependency on Biopython this way:

Note since it is a compile time requirement, the Biopython package itself will depend on NumPy, so you do not need to also declare that as an explicit dependency.

However, the package will not trigger the installation of optional packages such as ReportLab, MatPlotLib or NetworkX which parts of Biopython require.

So, if for example you are writing a Galaxy tool in Python using Biopython, MatPlotLib and SciPy you must explicitly declare them all as dependencies in your repository_dependencies.xml file, e.g.:

<?xml version="1.0"?>
<repositories description="Requires Biopython, MatPlotLib and SciPy.">
  <repository name="package_biopython_1_65" owner="biopython" />
  <repository name="package_matplotlib_1_2" owner="iuc" />
  <repository name="package_scipy_0_14" owner="iuc" />

And in your tool XML file(s) you would include:

  <requirement type="package" version="1.65">biopython</requirement>
  <requirement type="package" version="1.4">matplotlib</requirement>
  <requirement type="package" version="0.14">scipy</requirement>

This will work on the main Tool Shed refering to these packages (and similarly on the Test Tool Shed):

These additional dependencies are maintained by the IUC, or in full the "Intergalactic Utilities Commission" which is the tongue-in-cheek name of the Galaxy Tool Shed review team, see:

When looking for how to add further dependencies, search the Tool Shed for "Tool dependency definitions", and in particular look for those from the IUC:

In addition to the examples above, the following are of particular relevance:

For full details on how to define dependencies on the Galaxy Tool Shed, see