Running immuneML in the cloud

In this tutorial, we will show how immuneML can be run on Google Cloud and Amazon Web Services (AWS) using the existing Docker image.

Installing immuneML on Google Cloud

Here we will describe how immuneML can be set up on a virtual machine on Google Cloud, so that it is possible to log in to the machine and perform an analysis using immuneML.

To follow this tutorial, you will need an account on Google Cloud.

The following video shows how to set up immuneML once you have a Google Cloud account.

immuneML Docker image is available at Docker Hub.

Installing immuneML on AWS

Here we will describe how immuneML can be set up on AWS. We will use a single EC2 Linux instance and install immuneML using Docker. Once the instance is created, the process is the same as for any other installation of a Docker image on a Linux machine.

To install immuneML on AWS, follow these steps:

1. Set up to use Amazon EC2 instance (a single virtual machine) following the steps described here including signing up for AWS and creating a key pair and a security group for accessing virtual machines on AWS.

2. Launch an EC2 instance following this tutorial. When choosing the instance type, for running immuneML you will need at least 4GB of RAM and 30GB of storage (but more is recommended, depending on the analyses you plan to run).

3. Follow this tutorial to install Docker on Ubuntu (or choose an alternative operating system from the side menu if you choose a different operating system for your EC2 instance).

4. Once the Docker is installed, install immuneML Docker image and run a quickstart analysis:

docker run -it -v $(pwd):/data --name my_container milenapavlovic/immuneml immune-ml-quickstart /data/output/

This command will download the immuneML Docker image, start a Docker container called my_container from the downloaded image, run the quickstart, and store results in output directory in the current directory. For more information on immuneML and Docker, see the Setting up immuneML with Docker.