DeepVariant comes with scripts to build it on Ubuntu 14 and 16, with Ubuntu 16 recommended. It can likely be built and run on other unix-based systems with some minimal modifications to these scripts. One way to get access to a machine running Ubuntu is through a cloud computing platform like Google Cloud Engine.
First install the Google Cloud SDK,
because we will need to use its gsutil command to fetch some dependencies.
The build-prereq.sh command below will install a number of system packages to
fulfill DeepVariant's prerequisites (using apt-get and pip, invoked via sudo),
and will fetch some precompiled binaries (for Tensorflow and CLIF) from our
Google Cloud Storage bucket (using gsutil).
You could build Tensorflow and CLIF from source if you prefer; see their respective source distributions for details. Update the prereq script to use your installed versions if you do that, since we expect them to wind up in standard locations.
Run the following commands to install prerequisites, build the DeepVariant programs, and then run tests.
./build-prereq.sh
./build_and_test.shAt the end of the output of that last command, you should see a summary message like "Executed 55 out of 55 tests: 55 tests pass." along with the message "Target //deepvariant:binaries up-to-date:" followed by a list of the just-built deepvariant binaries.
The following command should be run on any machine on which you wish run DeepVariant, since there are runtime dependencies, such as Python packages like numpy and Tensorflow to be installed:
./run-prereq.shAdvanced users may want to edit the settings.sh file before building. It contains options for configuring TensorFlow, CUDA, GPU usage, etc.