Project Setup

In this project, we want to build a package that allows users to plot MAPQ statistics from an alignment file. A package is a collection of python files that contain functions, constants, and class definitions. These python files are usually referred to as modules. Modules are files that contains definitions that can be imported into other modules. A file’s name is the module name with the suffix .py.

In our package, we have the following items:

  • a collection of modules
  • build and installation scripts
  • scripts to run from the command line
  • unit tests
  • test data
  • documentation

Below, we describe these items in detail.

Basic Directory Structure

Let’s look at the structure of our project. Here, we have:

        - tests

In this example, PlotMAPQ is our package. Within this package, we have two modules: and PlotMAP. Once we install our package, we will be able to import these modules.

Note another file The file is required in package directories for python to treat directories as packages. files can be empty, but can also contain initialization code for the submodule it defines. Here, we only have two files, one in our package and one in our test directory. However, if your project has multiple subdirectories, you would have more.

The file is the script used for building and installing the package. This file uses setuptools, a library for facilitating the packaging of python projects. defines the package version, dependencies, metadata, and other build information. Let’s take a look at a part of our setup file:

import setuptools


    name="MAPQTools_myname",                                    # name of project
    install_requires=REQUIRED_PACKAGES,                         # all requirements used by this package
    version=this_version,                                       # project version, read from
    author="Author name",                                           # Author, shown on PyPI
    scripts = ['bin/plot-mapq'],                                # command line scripts installed
    author_email="",                                         # Author email
    description="Plots MAPQ distributions from BAM files",      # Short description of project
    long_description=long_description,                          # Long description, shown on PyPI
    long_description_content_type="text/markdown",              # Content type. Here, we used a markdown file.
    url="",   # github path
    packages=setuptools.find_packages(),                        # automatically finds packages in the current directory. You can also explictly list them.
    classifiers=[                                               # Classifiers give pip metadata about your project. See for a list of available classifiers.
        "Programming Language :: Python :: 3",
        "License :: OSI Approved :: Apache Software License",
        "Operating System :: OS Independent",
    python_requires='>=3.6',                                    # python version requirement

Note that we have also defined package requirements and a long description for PyPI in

Now, we will use our setup to build a distribution. The distribution consists of a tar archive of all the files needed to build and install the package.

python sdist

This will build a distribution under dist/ which can be uploaded to PyPI. For now, we want to play around with our project in the development mode, so we will install our project. In your project directory, install MAPQTools in editable mode:

pip install -e .

This will install our project in “editable” mode, meaning the package will be updated as we make modifications to our files.

Now that we have installed our package in editable mode, we can access the package from python:

>>> import PlotMAPQ


Note that we also include the scripts directive in This specifies a list of scripts that setuptools installs to be accessible by the command line. To add this functionality, we create a bin directory that contains scripts the user can run. We discuss this more in Command Line Scripts.

Unit tests

All good projects have unit tests. Luckily, there are a ton of great resources to add unit tests and run them regularily. We have added unit tests under PlotMAPQ/test. This directory contains unit tests and data required to run the unit tests. We discuss this in more detail in Unit tests and Continuous Integration.


Under the directory docs, we have a bunch of files that create the documentation you are reading right now! We will discuss this in detail later in Project Documentation.

Other Important Files

In addition to the files mentioned, we have some other files that are important to a project:

1. the README gives information for users on github, and is also used to provide information about your package on PyPI. In this example, we read the file in to provide a long description for our package. Our read me is a markdown (md) file.

  1. LICENSE: A license tells users of your package the terms under which they can use it. choosealicense provides information for picking out a license. In this tutorial, we use the Apache Foundation license, as it provides no restrictions for users.
  2. Makefile: A Makefile is completely optional, but makes it easier to build and develop projects. Here, we have added a makefile that contains objectives that allow developers to quickly run tests, develop, and push to PyPI.

Next Steps

For more information on package python projects, see python-packaging-tutorial. Next, we will take a look at our command line script.