Supplementary Code

This directory tree contains supplementary code supporting the analysis reported in the article "Pervasive prognostic signals in the cancer transcriptome or why association with outcome is not biologically informative".

Author: Gil Tomás gil.tomas@ulb.ac.be

URL: https://owncloud.ulb.ac.be/index.php/s/iAleeNNQ7adenTM

Contents

• Preliminaries and license
• Install
• Run the analysis
• Where things are

Preliminaries and license

The execution of this code requires a LINUX/UNIX environment with a working R (version>=3.1.2) and TeX installations. Its sole intent is to support the findings reported in the quoted article.

This file is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This file is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this file. If not, see http://www.gnu.org/licenses/.

Install

Prior requirements to the running of this software include a working R (version>=3.1.2) and TeX installations. Furthermore, R packages described in the file config/global.dcf are also expected to be found in your system. In addition, the CRAN R package ProjectTemplate (version>=0.6) and MicroarrayToolbox (available on http://github.com/gtms/MicroarrayToolbox) must also be installed.

An R script located on install/install-packages.R can be executed to fill in these requirements. On a bash command line, enter:

R CMD BATCH install/install-packages.R

Run the analysis

This project runs within the R ProjectTemplate framework for automated data analysis (http://projecttemplate.net).

Launch R at the root directory of the project, where this README.md file is located, or set the working directory with the setwd () command.

Then you need to run the following two lines of R code:

library ("ProjectTemplate")
load.project ()

Once the second line of code is evaluated, a series of automated tasks will be executed depending on the configurations declared in the config/global.dcf file. With the original configuration, these tasks include:

• Loading any R packages listed in the configuration file.
• Reading relevant datasets stored in data or cache.
• Pre-processing the data using the files in the munge directory.
• Executing the analysis of pre-processed data, yielding graphical output data.

Where things are

• Directories

  • Configuration files

    The analysis work-flow followed by ProjectTemplate is determined by the configuration flags found in the config/global.dcf file. Depending on the TRUE/FALSE status of these flags, the function load.project () may: load raw data into memory (flag data_loading); load pre-processed data into memory (flag cache_loading); pre-process raw data (flag munging); and load pre-determined libraries into memory (flag load_libraries). For instance, once the raw data has been initially pre-processed and cached, you may find it desirable to turn the munging flag off and the cache_loading flag on. This will allow for direct access to pre-processed data on your working environment once load.project () is executed on later R sessions.

  • Raw data

    Raw data can be found in the data directory. The csv directory contains the file studies.csv, which has information about all data-sets analyzed in this study. The rda directory contains all data-sets stored on disk as Rda files. The sigs directory contains biologically motivated gene expression signatures in Rda format, plus the 4722 MSigDB curated gene sets (collection v4.0, updated on May 31, 2013), as downloaded from http://www.broadinstitute.org/gsea/msigdb, in the gmt format.

  • Pre-processed data

    Pre-processed data, or cached data, can be found in the cache directory as Rda files. These are the output of the processing of raw data with scripts located in the munge directory.

  • Pre-processing scripts

    Pre-processing scripts are located in the munge directory.

  • The remaining directories should be self-explanatory.

• Output file types

  • *.Rout

    These are run logs, i.e. records of a particular computation run by a script. These include non-graphical intermediate results, values of the random number generator seeds, and software package versions.

  • *.pdf

    These are graphical outputs.

  • *.Rda

    These are binary files storing pre-processed intermediate results that can be further scrutinized should the user decide to fine-tune the analysis or push it in another direction.