Bioconductor Devel Docker image - DEPRECATED

Deprecation Notice : The bioconductor/bioconductor_full images have been deprecated in favor of the new bioconductor/bioconductor_docker images located at https://github.com/bioconductor/bioconductor_docker . These new images will have the same functionality as bioconductor_full, and users should find it easy to switch between the new set of containers. Please refer to the README.md of the new images for more information.

Bioconductor Docker image with full set of system dependencies so that all Bioconductor packages can be installed.

The Docker images have R and Bioconductor with different versions under each "branch" in git.

NOTE: Docker image for bioconductor_full:devel is in the master branch, and all the release branches will be under the branch RELEASE_X_Y.

Important Links:

Docker hub link for bioconductor_full

Github development link for bioconductor_full

Advantages of the bioconductor_full docker image

1. The bioconductor_full docker images can be used instead of installing complex dependencies needed for Bioconductor packages. The image comes with most of the dependencies installed.

2. Quick start up to start your analysis with all the Bioconductor packages if needed.

3. The image will be regularly updated to reflect the build system on Bioconductor. This is a very useful resource for maintainers who are actively developing their package to see if it works in tandem with the bioconductor ecosystem. It provides a local testing outlet for maintainers and developers.

Installation and quick start

This document assumes you have docker installed. Please check installation if you have more questions regarding this.

Quick start

• Start docker on your machine.

• On the command line, "pull" the bioconductor_full docker image with the correct tag. These images are hosted on docker hub under the Bioconductor organization page.

    docker pull bioconductor/bioconductor_full:devel
  

  or

    docker pull bioconductor/bioconductor_full:RELEASE_X_Y
  

• Once the image is available on your local machine, you can check to see if they are available.

    docker images
  

• To start using these images with RStudio, this will start the image under the 'rstudio' user

    docker run                                      \
        -e PASSWORD=your_password                   \
        -p 8787:8787                                \
        -v ~/host-site-library:/usr/local/lib/R/host-site-library \
        bioconductor/bioconductor_full:devel
  

• To start the image interactively using the bioc user

    docker run                                     \
        -it                                        \
        --user bioc                                \
        -v ~/host-site-library:/usr/local/lib/R/host-site-library \
        bioconductor/bioconductor_full:devel       \
        R
  

  NOTE: The path /usr/local/lib/R/host-site-library is mapped to .libPaths() in R. So, when R is started, all the libraries in the host directory, host-site-library are available to R. It is stored on your machine mounted from the volume you fill in place of host-site-library.

  These libraries will only work if they are pre-compiled with the same version of R that is in the docker image. To explain further, you would need the packages built with Bioconductor version '3.9' to work with R-3.6. Similarly, you'd need Bioconductor version '3.9' to work with R-3.6.z.

• To start the docker image in deamon mode, i.e, have the container running in the background use the -d option.

    sudo docker run -it                            \
        -d                                         \
        -v host-site-library:/usr/local/lib/R/host-site-library \
        --entrypoint /bin/bash                     \
        bioconductor/bioconductor_full:devel
  

  This will start the container in the background and keep it running. You may check the running processes using docker ps, and copy the container id.

    docker ps
  

  To attach to a container which is running in the background

    docker exec -it <container_id> bash
  

  NOTE: You can replace bash with R to start R directly in the container.

    docker exec -it <container_id> R
  

• To run multiple RStudio instances, use a different external port mapping (the first port in -p XXXX:YYYY) for each instance. Use standard shell commands (e.g., adding a & at the end of the first docker command) to place docker processes in the background. The 'devel' instance will be available at http://localhost:8787, and the release image at http://localhost:8788

    docker run                                      \
        -e PASSWORD=your_password                   \
        -p 8787:8787                                \
        bioconductor/bioconductor_full:devel
  
    docker run                                      \
        -e PASSWORD=your_password                   \
        -p 8788:8787                                \
        bioconductor/bioconductor_full:RELEASE_3_10
  

Issues in the devel image

Packages which will never install

1. xps

2. Rmpi

3. ccfindR

For more information on the issues for this image, please check issues.md

Singularity

Bioconductor provides singularity hub images. The files Singularity, Singularity.RELEASE_3_8, Singularity.RELEASE_3_9 allow the images to build on Singularity hub.

Find the singularity hub images, and download them by https://www.singularity-hub.org/collections/3154/usage

singularity pull shub://Bioconductor/bioconductor_full

Frequently Asked Questions (FAQ)

How do I use LaTeX inside the container?

Install the tinytex package (https://yihui.name/tinytex/) which has helpers for installing LaTeX functionality.

install.packages('tinytex')
tinytex::install_tinytex()