Description

RedRibbon: a rank-rank hypergeometric overlap library.

Installation

This procedure have been tested on debian/ubuntu but should work on any linux distribution.

Directly from Github

In R, just run

devtools::install_github("antpiron/RedRibbon")

Alternative installation method from the tar.gz released package

In R, download RedRibbon_0.3-1.tar.gz on github, and then the run

R CMD INSTALL RedRibbon_0.3-1.tar.gz

For advanced users only: C librairies if you plan to program in C (require root)

If you plan to program in C, standalone librairies can be installed.

We recommend the installation of GCC OpenMP support, the GOMP library. For Debian/Ubuntu:

sudo apt-get install libgomp1

If not installed, while the libraries will be totally functionnal, no parallel execution will be available.

Download official release of the ale and cRedRibbon C libraries.

Install ale C library

tar xvzf ale-1.1.tar.gz
cd ale-1.1
./configure
make
sudo make install

Install cRedRibbon C library

tar xvzf  credribbon-1.1.tar.gz
cd credribbon-1.1
./configure
make
sudo make install

If the two C librairies are installed, the R package installation process will automatically detect those and link to those.

Rermark: if the libraries are not automaticaly detected, sudo ldconfig /usr/local/lib/ might be necessary to run in order to update links and cache to the shared libraries.

Documentation

A R vignette with fully reproducible examples is available here. You can also open it from R with

vignette("RedRibbon")

Short summary,

library(RedRibbon)

## df is a data.frame with 3 columns: id, a, b
## 
##    id 	   a 	   b
## gene1 	-499 	-499
## gene2 	-498 	-498
## gene3 	-497 	-497
## gene4 	-496 	-496
## gene5 	-495 	-495
## gene6 	-494 	-494 
## ...

## Create RedRibbon object
rr <- RedRibbon(df, enrichment_mode="hyper-two-tailed")

## Run the overlap using evolutionnary algorithm,
## computing permutation adjusted p-value for the four quadrants
quad <- quadrants(rr, algorithm="ea", permutation=TRUE, whole=FALSE)

## Plots the results
ggRedRibbon(rr, quadrants=quad) + coord_fixed(ratio = 1, clip = "off")

## Get the down-down quandrant list of genes
df[quad$downdown$positions,]

Citation

Please cite the biorxiv pre-print,

RedRibbon: A new rank-rank hypergeometric overlap pipeline to compare gene and transcript expression signatures
Anthony Piron, Florian Szymczak, Maria Inês Alvelos, Matthieu Defrance, Tom Lenaerts, Décio L. Eizirik, Miriam Cnop
bioRxiv 2022.08.31.505818; doi: https://doi.org/10.1101/2022.08.31.505818