Public scRNA-Seq Datasets

This repository is used to build scater objects and reports (in continuous integration manner) for various publicly available scRNA-Seq datasets used by our group. This pipeline is implemented using docker containers and cloud computing. The resulting website is available here. Below are some notes on the pipeline setup.

Website

The website is generated using MkDocs generator. Links to S3 storage and data annotations are added manually. If you are creating a pull request and adding new data please add its annotations to one of the files in the website folder.

S3 storage

To list files on the S3 storage please use this link.

Instance setup

AWS

1. Launch Amazon Linux EC2 instance.
2. Using security groups add access to the instance on port 8080.
3. Connect to instance and install Jenkins.
4. Add permission for Jenkins to run Docker:

sudo usermod -aG docker jenkins

Hard reboot your instance after that. Now Jenkins can run docker images.

5. Install s3cmd utility to be able to upload data to the S3 storage:

sudo apt-get install s3cmd

6. In Jenkins Export S3 key ID, secret key and region as environmental variables. Use secret text option provided by Jenkins. Some details available here.

7. File listing can be setup on AWS S3 bucket using this plugin.

OpenStack Cloud (Sanger)

1. Launch Ubuntu Trusty instance (j1.large flavour)
2. Add the instance to the default, cloudforms_icmp_in, cloudforms_ssh_in cloudforms_web_in security groups.
3. Create additional security group: TCP with port 8080 (this is needed for Jenkins) and add your instance to this group.
4. Associate a floating IP (FLOATING_IP) number with your instance.
5. Login to instance:

ssh -i ~/.ssh/your_key.pem [email protected]_IP

6. In the instance install Jenkins.

To setup Jenkins after installation go to http://FLOATING_IP:8080

7. In the instance install docker.

8. Resolve docker network issues (Sanger OpenStack problem only):

sudo bash -c "echo '{ \"bip\": \"10.10.0.1/16\", \"mtu\": 1400 }' > /etc/docker/daemon.json"

9. Add permission for Jenkins to run Docker:

sudo usermod -aG docker jenkins

Hard reboot your instance after that. Now Jenkins can run docker images.

10. Install s3cmd utility to be able to upload data to the S3 storage:

sudo apt-get install s3cmd

11. In Jenkins Export S3 key ID, secret key and region as environmental variables. Use secret text option provided by Jenkins. Some details available here.

Jenkins build

# build and deploy
sh deploy.sh $WORKSPACE

AWS Calculator

To calculate how much you can spend on AWS one can us the AWS Calculator.