This project outlines using a Docker image to run RStudio server with common population genetics tools and packages included. It is build from the RStudio server image from the Rocker Project.
Currently the image contains the following tools:
- RStudio 4.3.2
- BLAST+
- ClustalW2
- ClustalO
- KaKs_Calculator
The image also contains the following R packages (among many standard ones):
- rmarkdown
- tidyverse
- devtools
- workflowr
- BiocManager
- rdiamond
- metablastr
- orthologr
- ggmsa
It is easy to add packages from RStudio Server in the container, preserved for future use in the same container as long as it's not removed.
This image is based on Dave Tang's work highlighted in his blog post Running RStudio Server with Docker.
- Docker for your platform (on Windows and Mac use Docker Desktop)
This image can be run as a container on Linux (x86_64/amd64), Apple silicon, Apple Intel or Windows 11 machines with the proper setup of Docker.
Important: run the following commands from a project folder to which you'd like access from the container.
To try this project on Linux ARM64 (e.g., on Apple Silicon+macOS) or Linux AMD64 (e.g., on Linux x86_64 machines), use the following command and pre-built image.
docker run --name rstudio_server -d -p 8888:8787 -v $PWD:/home/rstudio -e PASSWORD=password -t rheartpython/rstudio-pop-genetics
To try on Intel-based Macs use the following command.
docker run --platform=linux/amd64 --name rstudio_server -d -p 8888:8787 -v $PWD:/home/rstudio -e PASSWORD=password -t rheartpython/rstudio-pop-genetics
To run on Windows, the command in Command Prompt will be similar except for the -v $PWD:/home/rstudio part which will use Windows environment variable syntax instead (e.g., -v %CD%:/home/rstudio), thus the Windows 11 command is as follows in Command Prompt.
docker run --name rstudio_server -d -p 8888:8787 -v %CD%:/home/rstudio -e PASSWORD=password -t rheartpython/rstudio-pop-genetics
Log in to RStudio (at http://localhost:8888/) with username "rstudio" and password "password" and start having fun.
To build or modify and build your own image, keep reading.
The following instructions are for building on Apple silicon (tested on an M3).
Replace <username> with a local name of your choosing or your Dockerhub username if pushing to Dockerhub.
Run the following to build for Apple silicon.
docker build --platform=linux/arm64 --rm=true -t <username>/rstudio-pop-genetics:arm64 .
Run the following commands to build cross-platform (linux amd64 and linux arm64), pushing to a container registry, an image that can be run on an amd64 Linux machine or Apple silicon.
docker login
docker buildx create --use
docker buildx build --platform linux/amd64,linux/arm64 --rm=true -t <username>/rstudio-pop-genetics --push .
Note, the base image does not have a build for darwin/amd64, but can still work when using the flag --platform=linux/amd64 on Intel-based Macs.
The local folder where the docker run command is run will be shared with the container. Replace <username> with a local name or your Dockerhub username if you have pushed to Dockerhub.
Run the following command in your local folder of choice (usually where your project files reside) to start the container on Apple silicon (arm64) platforms.
docker run --name rstudio_server -d -p 8888:8787 -v $PWD:/home/rstudio -e PASSWORD=password -t <username>/rstudio-pop-genetics
Refer to the Quickstart guide above for other methods of running the docker container.
Log in to RStudio (at http://localhost:8888/) with username "rstudio" and password "password" and start having fun.
To stop the container (this will keep the state of packages and files):
docker stop rstudio_server
To remove the container from Docker (note, this will not save any installed packages, however the host files loaded into RStudio will be preserved in their latest state as long as the volume was mounted/shared):
docker rm rstudio_server
To start a bash session for the running container:
docker exec -it rstudio_server /bin/bash
In the bash session:
service rstudio-server restart
exit