R is a programming language, with many useful statistical functions included. One can gain additional functionality for microarray data analysis by adding the Bioconductor package to it. This is a set of biologically useful statistical tools, which, for example, allow loading of many of the common microarray formats of data. One of the good things about R/Bioconductor is that they are both free. There is also a wealth of information on the internet, including vignettes on how to use each function. Many papers have been published where R/Bioconductor have been used to analyse the microarray data.

This set of instructions are for installing R/Bioconductor on Windows XP. You can also get R and BioConductor on Mac OS X, Linux and other Unix systems.

Installing R

First we need to download a copy of R from The R Project. They have many "mirror" sites around the world that you can download from, it is fastest to use a local British site such as:

http://www.stats.bris.ac.uk/R/bin/windows/base/

Choose the 'Setup program' link (about 28 megabytes), it should be called something like R-2.4.1-win32.exe (this is the current version as of 08/01/2005). You will need to run this file, but you can choose to save it first if you desire.

Running the file and following the prompts should install R.

Installing Packages

BioConductor is exceptional (see below) - installing most packages is very easy.

Normal packages can be installed from within R by selecting the "Packages" menu, and clicking on from the "Install Package(s)..." entry.

R will then try and access the internet (give it permission if your firewall complains) and ask you to choose a CRAN mirror site. Choose anything in the UK as these should be fastest.

Then a long alphabetical list of available packages will be shown. Find the one you want, and click "OK" to install it.

Its very easy.

Loading Packages

Once installed, you must still load a package before you can use it. To load a package you can type the command library(packagename), or use the menu "Packages", "Load Packages..." and select from the list of installed packages.

Installing BioConductor

This takes a long time, and you may not need BioConductor for your assignments

To install BioConductor, we first need to start R. You should find it on the start menu; the exact location depends on your choices during installation.

Once R is open we need to add the functionality of the Bioconductor package. This is as simple as typing in the following commands (inside R), pressing enter after each line (make sure you get the capitalization correct, you can use copy and paste if you prefer):

source("http://www.bioconductor.org/biocLite.R")
biocLite()

That will install the basic functionality. You may prefer to do a full installation instead:

source('http://www.bioconductor.org/getBioC.R')
getBioC()

This may take a little while to run (and the full installation takes a lot longer), but you only have to do it once. Various messages will flash up on the screen tell you about what is being downloaded. Ignore warnings such as 'unable to access index for repository' etc.

(This is probably an ideal time to get yourself, and whoever is running the class, a nice cup of tea!)

You have now installed R with Bioconductor, and can close the R window.

The R Project

 

BioConductor