DNAnexus Platform SDK

The DNAnexus Platform SDK (a.k.a. dx-toolkit) includes the dx command-line client, tools for building and debugging apps, other utilities for working with DNA data on the DNAnexus platform, and Python, Java, C++ and R bindings for working with the DNAnexus platform.

Our SDK is released under the Apache v2.0 license.

Note: If you plan to do most of your interaction with the platform through the web UI, and will use the command-line only to upload files, you probably do not need to download the SDK. The Upload Agent (discussed below) is our specialized tool for uploading to the platform and may be all that you need from this page.

Installation of Python SDK and Command Line Tools

Versions of Mac OS X 10.8 and higher come with Python 2.7 preinstalled. However, DNAnexus is deprecating usage of Python 2.7 and it is recommended that you check your version of Python to ensure you are using Python 3 (links to Windows and Linux downloads as well) instead before moving onto the next steps.

The Python components of the SDK (dxpy bindings and command-line tools including dx) are distributed through PyPI. Use pip3 to download and install the dxpy package through the following command:

pip3 install dxpy

To enable tab completion after installing with pip3, run the following command or add it to your .bashrc

eval "$(register-python-argcomplete dx|sed 's/-o default//')"

If you use later version of MacOS with Z shell (zsh), enable tab completion by running the following command or adding it to your .zshrc

autoload -Uz compinit && compinit
autoload bashcompinit && bashcompinit
eval "$(register-python-argcomplete dx|sed 's/-o default//')"

Congratulations! You are now ready to use the dx command-line tool and the dxpy Python module.

What's next:

• The Command Line Quickstart will get you started on accessing the platform from the command-line.

• The Developer section contains comprehensive information on how to use the SDK to interact with and write your own applications for the DNAnexus Platform.

Proxy and TLS Settings

To direct dx-toolkit to a proxy, please set the HTTP_PROXY environment variable to something like export HTTP_PROXY=hostname:port. HTTPS_PROXY is also supported.

By default, dx-toolkit uses certificates installed on the system to create secure connections. If your system requires an additional TLS certificate and the dx-toolkit doesn't appear to be using a certificate installed on your system, there are two options in order of preference. First, set the DX_CA_CERT environment variable to the path of the PEM-encoded TLS certificate file required by your parent organization. You can also connect insecurely by avoiding certificate verification all together by setting DX_CA_CERT=NOVERIFY.

For Windows users, the way environments are customized can vary enough to where the settings above may still cause issues. Please contact [email protected] if you are having trouble connecting through a proxy.

Installation of Java, R and C++ SDK from tarball

You can download the SDK for Java, R, C++, and Python2.7 built for your platform:

Install

Linux and Mac

1. Unpack the SDK:

   $ tar -xzf dx-toolkit*.tar.gz

   Note: The tarballs include Python 2.7 version of dxpy bindings and dx command, which will be supported until March 1, 2021.

2. Initialize your environment as follows:

   $ <PATH_TO_DX_TOOLKIT> dx-toolkit/environment

Windows (Windows 10 and later versions are supported)

1. Run dx-toolkit*.exe to start the installation. Systems with Windows Defender enabled may present a warning before launching the dx-toolkit installer. If you see this warning window, click "More Info", and choose "Run Anyway". The sources for the Windows installer are here.

2. After the installation is completed, find and click on the DNAnexus CLI shortcut on the Windows Start Menu.

If your organization does not allow installation on Windows systems without the installer package's signature, please get in contact with the support team at [email protected] for possible alternative installation method.

Congratulations! You are now ready to use the Platform SDK.

Upgrading the tarball installation

The dx upgrade command will upgrade your DNAnexus SDK to the latest recommended version. The dx upgrade command can also revert your dx-toolkit to a previous version.

Calling dx upgrade without any arguments will upgrade your DNAnexus SDK to the most recent version. Alternatively, you can provide dx upgrade with a version of the DNAnexus SDK that you would prefer to have installed on your local system.

Upgrading to the latest version:

$ dx upgrade
Upgrading to v0.306.0
Downloading dx-toolkit-v0.312.0-osx.tar.gz (using curl -O -L)...
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 9874k  100 9874k    0     0  3279k      0  0:00:03  0:00:03 --:--:-- 3279k
Unpacking dx-toolkit-v0.312.0-osx.tar.gz...
Backing up version v0.227.1...
Moving version v0.306.0 into place...
upgrade.sh: Updated to version v0.306.0. Previous version saved in /Users/user-amy/dnanexus/dx-toolkit/build/v0.227.1.
Please close this terminal, open a new terminal, and source the environment file in the dx-toolkit folder

Upgrading to a previous version:

$ dx upgrade v0.225.0

Ubuntu APT Repository

If you are running Ubuntu 16.04, see Ubuntu Repository for instructions on how to obtain the DNAnexus SDK via APT.

Building from source

Download either of the following source packages:

See the Readme.md file in the archive for build-time dependencies and further installation instructions (or view on Github).

Upload Agent

The Upload Agent is a command-line tool for uploading files to the DNAnexus Platform. It is a fast and convenient alternative to using the web interface for uploading files.

Features

• Fast Parallel Uploads: Data is uploaded using multiple parallel connections for faster upload speeds.

• Automatic Compression: Text files (such as FASTA, FASTQ, qseq, and others) can be automatically compressed before upload, for better efficiency.

• Automated Uploads: No user interaction is required; the client can be used to automate uploads of entire sequencing runs.

• Resumable Uploads: Previously interrupted uploads can be automatically resumed.

• Progress Reporting: You can monitor individual file upload progress and also average/instantaneous transfer rates.

Downloads

The Upload Agent is available pre-built for the three main flavors of operating systems:

Install

Unpack and run the downloaded file. You should then see a help message.

• 64-bit Linux:

  $ tar -xzf dnanexus-upload-agent-*-linux.tar.gz
  $ cd dnanexus-upload-agent-*-linux
  $ ./ua

• Mac OS X

  $ unzip dnanexus-upload-agent-*-osx.zip
  $ cd dnanexus-upload-agent-*-osx
  $ ./ua

• Windows

  • Unzip the file using your favorite utility

  • Run "ua.exe"

This program requires that you have an API key. Please create an API key before proceeding.

Download Agent

The Download Agent is a command-line tool for downloading large quantities of files from DNAnexus to a local environment. It is designed to resiliently manage downloads over potentially long periods of time.

Features and Usage

For a detailed description of Download Agent features and usage, please review the README file on its Github location.

Downloads

A list of Download Agent releases is provided on the Releases section of its Github location. The Docker repository additionally provides access to experimental builds corresponding to commits to the master branch.

dxCompiler

dxCompiler is used to generate a DNAnexus workflow from a workflow written in Workflow Description Language (WDL) . dxCompiler is a JAR file that can be run with Java 8 or later, and can be downloaded from the releases section of the DNAnexus dxCompiler Github repository. dxCompiler is also released as a public docker container that includes the Java dependency. dxCompiler README documents the setup instructions and sample workflows.

Before using dxCompiler, ensure you have the following dependencies installed:

1. dxcommand-line tool is installed and initialized withdx login

2. • ​Java SE Development Kit 8 (or above) is required to invoke dxCompiler with java -jar dxCompiler.jar compile workflow_file.wdl

   • Alternatively, docker command-line tool is required to invoke the dxCompiler inside the docker container using the run-dxcompiler-docker script