Module 3 Bioconductor

  • Learn about Bioconductor, a repository for R packages
  • Learn about the BiocManager package and how to use it to install R packages hosted by Bioconductor

3.1 What is Bioconductor?

Bioconductor is a repository of R packages. All Bioconductor packages are regularly built and tested to ensure that they are functioning correctly and meet certain standards.

For every new release of R a snapshot of the entire Bioconductor repository is stored and retained in order to make analysis reproducible at the time it was run.

It is essential therefore that you keep up-to-date with the R version, otherwise you will not be able to use the latest version of packages in the Bioconductor repository.

The BiocManager package (link) contains a small number of functions to install R packages from Bioconductor (and other repositories such as CRAN) and check that the versions are compatible with the version of R you are using.

3.2 Installing Bioconductor packages

Using BiocManager to install the packages needed for this course will ensure that you are using the most recent release, and can also install the appropriate versions of all needed dependencies.

Install packages

To install the packages for this course copy and paste the code below into the command line and press Enter:

BiocManager::install(
    c('struct','structToolbox','metabolomicsWorkbenchR'),
    dependencies='suggests')

The dependencies input ensures that we install all suggested packages required to use the ones we are installing. There will be quite a few of them!

You can check that the packages are installed and working by running the following code:

library('struct')
library('structToolbox')
library('metabolomicsWorkbenchR')

You might receive some (sometimes cryptic!) messages after activating the packages, but as long as you receive no error messages, then you are ready to go.

3.3 Additional packages

In addition to the struct and structToolbox packages it is useful to load the ggplot2 package. It will allow you to modify e.g. chart titles quickly and easily. It should have been installed already as a dependency of structToolbox, so you can just activate it for the current session.

In addition you will need the following packages:

library(ggplot2)

In addition, you will also need make sure the following R packages are installed:

  • dplyr
  • xcms
  • MSnbase
  • openxlsx
  • RColorBrewer
  • ggplot2

We strongly recommend you continue to use BiocManager:install rather than install.packages to install additional packages if possible, even if they are not a Bioconductor package (e.g. CRAN packages). This will ensure that a version compatible with your R and Bioconductor versions is used.