DNAnexus Platform SDK
The DNAnexus Platform SDK (a.k.a.
dx-toolkit) includes the dx command-line client, tools for building and debugging apps, and other utilities for working with DNA data on the platform. Our SDK is released under the Apache v2.0 license.
If you plan to do most of your interaction with the platform via the website, 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.
You can download the SDK built for your platform:
See the section Alternative downloads below for alternative ways to install the SDK.
- Unpack the SDK:
tar -xzf dx-toolkit*.tar.gz
Note: Python 2.7 is required. Command-line tools in the SDK will not work with Python 2.6 or Python 3.
- Initialize your environment as follows:
Congratulations! You are now ready to use the Platform SDK.
- The Command-Line Quickstart will get you started on accessing the platform from the command-line.
- The Developer Portal contains comprehensive information on how to use the SDK to interact with and write your own applications for the DNAnexus Platform.
Building from source
Download either of the following source packages:
Readme.md file in the archive for build-time dependencies and further installation instructions (or view on Github).
The Python components of the SDK (
dxpy bindings and most of the command-line tools) are distributed through PyPI. Use pip to download and install the
Note: OS X 10.9 users with Python < 2.7.6 from python.org or pythonbrew may encounter Python issue 18458
when using the Python interpreter interactively. Upgrade to Python 2.7.6 or greater to resolve the issue.
Ubuntu APT repository
If you are running Ubuntu 12.04 or later, see Ubuntu Repository for instructions on
how to obtain the DNAnexus SDK via APT.
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.
- 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
- 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
- Progress Reporting: You can monitor individual file upload
progress and also average/instantaneous transfer rates.
The Upload Agent is available pre-built for the three main flavors of
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
- Mac OS X
$ unzip dnanexus-upload-agent-*-osx.zip
$ cd dnanexus-upload-agent-*-osx
- Unzip the file using your favorite utility
- Run "ua.exe"
Note for Windows 7/2008 users:
If you receive an error message saying: 'schannel: failed to setup extended errors', please install this hotfix
This program requires that you have an API key. Please create an
API key before proceeding.
See the Upload Agent Guide for more info about how
to use the tool.
dxR: R Bindings Package
There is a basic R bindings library called
dxR with wrapper functions for making API calls. Additional higher-level capabilities are under development.
Note: This package is currently under development and still in flux. Releases may introduce some incompatibilities with previous versions, though we will try to avoid this as much as possible and document when it does occur.
You can download the source package or a precompiled binary for Mac or Windows below.
- Install any missing software dependencies. You will need R, as well as the packages
RJSONIO. You can install these packages from within R with the following commands.
- You should now be able to install the
dxR package. You can use the R graphical interface to select and install the downloaded file. Or, you can run the following command using the path to the file you downloaded.
Alternatively, you can also run the following command from the command-line (outside of R) to install the package.
You can use the
> install.packages("dxR_*", repos=NULL)
-l flag to specify a local directory in which to
install the package if you do not have system permissions to install
it for all users. If taking such an approach, you should set your
R_LIBS environment variable appropriately if you have not already.
R CMD INSTALL dxR_* -l $R_LIBS