CSC8309 -- Gene Expression and Proteomics

Why R?

This practical uses the R statistical language to perform the majority of the analyses. I've decided to use R for a variety of reasons:

• You should have used it before.
• It is a free environment. Previously, we used GeneSpring which is quite nice, but the licenses would run out shortly after the module finished.
• R comes with extensive support for microarray and bioinformatics analysis in the shape of Bioconductor.
• There is good support available on the web for it's use.
• It's easy to automate — you just run your code multiple times.

Why not R?

Like much bioinformatics, R and bioconductor can be a bit tricky to use. The documentation is not always the best, and tends to get a bit out of date. You will need to use your knowlege and experience to cope if something goes wrong. In some cases, parts of the practical are broken and are not fixable. This is also the way; things change, instructions go out of date, methods that should work or have worked no longer do; it's something that you have to get used to. For this reason, I have left these sections in the practical but commented out the code.

R can also be reasonbly slow. If your machine gets a big laggy while it is running the code, be patient.

Let me repeat. R CAN BE SLOW. If nothing appears to be happening, then check the process information on your machine. Probably, it's doing lots of terribly hard sums.

Presentation

The R code snippets will be presented like so:

## install the basic bioconductor packages
#source( "http://bioconductor.org/biocLite.R")
#biocLite()
#biocLite('soybeanprobe')


## ensure installation before carrying on.

The code has been syntax highlighted which should make it a little easier to see what is going on. It's just plain text, though. You should be able to cut and paste it okay.

For each code snippet there is also a "(Complete File)" link. This is an .r file which contains the snippet (at the end) and all of the rest of the R which is necessary to run it — most of the snippets require that you have run other parts first. If you run this file in a new R environment, it should run, without errors and evaluate the snippet at the end. Secondly, there is a "(Rout)" link. This takes you to the output of the R running the complete file, not just the snippet that you see.

You should try and follow the practical through from one part to the next, but if it something goes wrong, the complete file should get you started again.