Gene-Specific Phenotype EstimatoR
Switch branches/tags
Nothing to show
Clone or download
Pull request Compare This branch is even with cbg-ethz:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
R
data
inst
man
vignettes
.gitignore
DESCRIPTION
NAMESPACE
README.md

README.md

gespeR: Gene-Specific Phenotype EstimatoR

gespeR is a novel model to estimate gene-specific phenotypes from off-target confounded RNAi screens. The observed phenotype for each siRNA is modeled as a the weighted linear combination of gene-specific phenotypes from the on- and all off-target genes. This deconvolution approach yields highly reproducible phenotypes, essential for unbiased analyses of siRNA screening data.

Reference:

Fabian Schmich, Ewa Szczurek, Saskia Kreibich, Sabrina Dilling, Daniel Andritschke, Alain Casanova, Shyan Huey Low, Simone Eicher, Simone Muntwiler, Mario Emmenlauer, Pauli Ramo, Raquel Conde-Alvarez, Christian von Mering, Wolf-Dietrich Hardt, Christoph Dehio and Niko Beerenwinkel.
gespeR: a statistical model for deconvoluting off-target-confounded RNA interference screens
Genome Biology, 2015.

Installation

gespeR is hosted on GitHub and available through Bioconductor. The package is released under the GNU General Public License (GPL) version 3 and includes examples and a vignette.

Data

In addition to the phenotypic readout, gespeR requires siRNA-to-gene target relation matrices, quantifying how strongly each siRNA downregulates transcript genes via on- and off-targeting. These matrices can be computed with miRNA target prediction tools, such as for instance TargetScan.

Target Relation Matrices

Below, we provide pre-computed matrices for all libraries used in Schmich et al., 2015. Wrapper scripts to run TargetScan in batch mode for the prediction of siRNA-to-gene target relation matrices are also available on GitHub, including a README with step-by-step instructions. All pre-computed siRNA-to-gene target relation matrices are stored in .rds files using R's serialization interface for single objects. Load the data into R by using the function readRDS(). Note that loading target relation matrices can require up to 5GB of RAM.

Pathogen Infection Screen Phenotypes

High-content, image based phenotypes for pathogen infection RNAi screens from the InfectX consortium are hosted at PubChem.

Usage

Step-by-step instructions demonstrating how to download, pre-process and deconvolute pathogen infection screen phnotypes is available in form of an R/Vignette.

Contributions

###Contact

Fabian Schmich
fabian.schmich (at) bsse.ethz.ch