Instructions for running in Docker

Introduction

Below are details of how to get the CollateralVulnerability2016 R package running inside a docker container. This is useful because it removes any machine or platform-specific dependencies or issues that may cause problems when running the pipeline and makes the pipeline more portable.

Step-by-step instructions

Make a VM with enough RAM and CPU -

    docker-machine create --driver virtualbox --virtualbox-cpu-count 4  --virtualbox-memory "12024" bigvm     

Set big VM as default:

    eval $(docker-machine env bigvm)     

Launch a new docker container from bioconductor -

    docker run -d -p 8787:8787 bioconductor/release_core     

Get the IP address of the docker machine:

    docker-machine ip bigvm     

Can then log into rstudio on VM as rstudio/rstudio:

    http://ip-address:8787/      

Launch a bash session on a running container:

    docker exec -it <container-id> bash     

Launch R and install packages

    source("https://bioconductor.org/biocLite.R")    
    biocLite()    
    biocLite(c('GOSemSim', 'biomaRt', 'ensembldb', 'EnsDb.Hsapiens.v79', 'RTCGA'))     
    install.packages(c('dplyr', 'tidyr', 'shiny', 'BiSEp', 'devtools', 'readr', 'DT', 'roxygen2'))     
    devtools::install_github('chapmandu2/CollateralVulnerability2016')     

Come out of R into bash to install ruby and intermine gem

    apt-get install ruby-full     
    gem install intermine  
   exit

You will now have a virtual machine running a docker container that can run the CV pipeline successfully. Now make an image so that next time we don't have to go through all of the installation:

    docker stop
    docker commit <installed-container-id> <installed-image-name>

Next time we can launch a container with the image as below, we also open an additional port for shiny to use:

    docker run -d -p 8787:8787 -p 6793:6793  <installed-image-name>

When we get to the point of the R script where we launch the shiny app, we then just need to change the following options:

    options(shiny.port = 6793)
    options(shiny.host = '0.0.0.0')

And we can then view the shiny app in the host machine browser as follows:

    http://ip-address:6793/

See Docker website for more more details...

To do:

• Implement as dockerfile
• Share volumes