Introduction

Shiny application of BioInstaller has been supported since v0.3.5. A pre-built Docker image can help you to start all configured BioInstaller services in one-click, which includes the R (v3.4.4), Opencpu server (v??), Shiny server (v??), and BioInstaller (v0.3.5).

How to deploy the BioInstaller using Docker

You need install the Docker program for futher opreations. It is required for running any Docker image including BioInstaller.

docker pull bioinstaller/bioinstaller
docker run -it -p 80:80 -p 8004:8004 -v /tmp/download:/tmp/download bioinstaller/bioinstaller

When the docker container of BioInstaller was created, you can access the three types of web service in the browser.

  • localhost/ocpu/ Opencpu server
  • localhost/shiny/BioInstaller Shiny server
  • localhost/rstudio/ Rstudio server (opencpu/opencpu)

Opencpu server

#Show all items supported by BioInstaller
curl http://localhost/ocpu/library/BioInstaller/R/install.bioinfo -d \
"show.all.names=TRUE" X POST
#Show all versions of Bwa
curl http://localhost/ocpu/library/BioInstaller/R/install.bioinfo \
-d "name='bwa', show.all.versions=TRUE" -X POST

#Download and install latest Bwa
curl http://localhost/ocpu/library/BioInstaller/R/install.bioinfo \
-d "name='bwa', destdir='/opt/aliner/bwa'" -X POST
# Download 1000 Genome Project annotation database
curl http://localhost:5656/ocpu/library/BioInstaller/R/install.bioinfo \
-d "name='db_annovar_1000g', destdir = '/opt/annovardb', \
extra.list=list(buildver='hg19')" -X POST

# Get the character key and retrieve the output
# Get the JSON format value of returned output
curl http://localhost/ocpu/tmp/{key}/R/.val/json
# Get the text format value of returned output
curl http://localhost/ocpu/tmp/{key}/R/.val/text
# Get the function ‘print’ output of returned value
curl http://localhost/ocpu/tmp/{key}/R/.val/print

Shiny server

Open browser and input the URL “localhost/shiny/BioInstaller”. It will activate a Shiny session and start the BioInstaller Shiny application.

Rstudio server

The Rstudio server was provided for users to connect the docker container just like in the desktop version of Rstudio.

How to start the standalone Shiny service of BioInstaller

If the users can not to run a docker image, the standalone shiny service of BioInstaller is an another way to start the service.

# If the BioInstaller R package has not been installed, you
# need execute the followed commands:
# install.packages('BioInstaller') It is better to start the
# worker in another background processor
BioInstaller::set_shiny_workers(3)
BioInstaller::web(auto_create = TRUE)

To reduce the dependence of BioInstaller, we do not add the Shiny application modules dependent packages in the DESCRIPTION file. The function ‘BioInstaller::web’ will check the required dependences of Shiny application and install them if the dependence not be installed.