Workspace for data science projects and NGS pipelines by Konrad Stawiski. Contains everything you can imagine. Can connect to tailscale network to bypass firewalls. Also contains OmicSelector preinstalled.
This is the Docker image with VSCode, RStudio, Jupyter Notebook and file manager running. Great for development.
docker pull kstawiski/seq-pipeline
docker run --rm -d --name kgs24-work --hostname $(hostname)-work -v $(pwd):/work/ -v /home/$USER/:/home/$USER/ -p 26969:80 --privileged -v /var/run/docker.sock:/var/run/docker.sock kstawiski/seq-pipeline
Command above mounts your working directory as /work
. Just go to http://localhost:26969
to start work or setup tunnel as below.
Note that when using Docker in Docker, for example by running the WDL pipeline via Cromwell, the path /work
will not work. You can make a symlink on host device or bind the exact same path to the docker container.
For NVIDIA CUDA support use kstawiski/seq-pipeline-gpu
image:
docker pull kstawiski/seq-pipeline-gpu
docker run --rm -d --gpus all --name kgs24-work --hostname $(hostname)-work -v $(pwd):/work/ -p 26969:80 --privileged -v /home/$USER/:/home/$USER/ -v /var/run/docker.sock:/var/run/docker.sock kstawiski/seq-pipeline-gpu
You need your free Tailscale network ID XXX
, than run:
docker exec kgs24-work tailscale up --advertise-exit-node --accept-routes --authkey=XXX
Authkey setup - https://login.tailscale.com/admin/settings/keys
Alternatively, you can use ngrok, e.g. /ngrok http 8080 --log=stdout > /ngrok.log &
(may require ngrok setup, ngrok is located in /ngrok
).
Alternatively, you can set up directory for projects and operate using delivered pipelines and scripts:
git clone https://github.com/kstawiski/seq-pipeline.git .
./work.sh project_name
- Whole Exome Sequencing (tumor only somatic mutations) - TO DO
- RStudio - https://www.rstudio.com/
- Jupyter Notebook - https://jupyter.org/
- VSCode - https://code.visualstudio.com/, https://github.com/coder/code-server
- Terminal is using https://github.com/butlerx/wetty
- Resources monitor uses https://github.com/DanielnetoDotCom/ServerMonitor
- Notes use https://github.com/artyuum/simple-note