Data:
- Motifs - Individually or tar.gz
- Data - Individually (with description) or tar.gz
- R Scripts - tar.gz
- Everything - tar.gz (Recommended - only 12mb!)
How to repeat the experiment:
Firstly, you will need a modern version of R, at least version 2.12.0. You will also need the following modules installed:
- preprocessCore (from BioConductor)
- gplots
- boot
- mixtools
Then, download and untar the everything.tar.gz file available above, change into the created directory, and run perform_main_experiments.R . That will create:
- Plots of t-statistics for the PCA regression and normal regression for each model.
- Weighted loading of each PC.
- A plot comparing the overall adjusted R-squared of each model.
- Results will also be saved in rData files for examination later.
For example, running the regression on a Linux box is shown below:
rob@host:~/scratch$ tar zxf everything.tar.gz
rob@host:~/scratch$ cd McLeay_and_Bailey_2011_expression-0.1/
rob@host:~/scratch/McLeay_and_Bailey_2011_expression-0.1$ R < perform_main_experiments.R
> #
> #
> # OK, here we perform the main experiments for the paper.
> #
> #
> source('regression_lib.R')
....
Plots will be saved as PDF files. If you have any further questions, please feel free to contact Robert.