eREVEALER

eREVEALER (enhanced REpeated eValuation of variablEs conditionAL Entropy and Redundancy) is a powerful method for identifying groups of genomic alterations that, together, associate with functional activation, gene dependency, or drug response profiles. By combining these alterations, eREVEALER explains a larger fraction of samples displaying functional target activation or sensitivity than any individual alteration considered in isolation. eREVEALER extends the capabilities of the original REVEALER by handling larger sample sizes with significantly higher speed.

Preprint is avaiable here

Overview

eREVEALER consists of two main components: REVEALER preprocess and REVEALER run.

• REVEALER preprocess: If you start with a MAF file or a GCT file that needs further filtering, run REVEALER preprocess first and use its output as the input for REVEALER run.
• REVEALER run: If you have a ready-to-use GCT format matrix, you can directly run REVEALER run.

For detailed documentation regarding each parameter and workflow, refer to the individual documentation for REVEALER_preprocess and REVEALER.

Installation

Python version prerequisite

Please use Python version >= 3.7 and < 3.10

Create Conda environment

conda create -n revealer python==3.9

Install via pip

eREVEALER can be used in the command line, Jupyter Notebook, and GenePattern. To use eREVEALER in the command line or Jupyter Notebook, install it via pip:

pip install revealer

Install via cloning the repository

Alternatively, you can install eREVEALER by cloning the repository and running the setup script.

1. Clone the repository:

   git clone https://github.com/yoshihiko1218/eREVEALER.git
   cd eREVEALER

2. Install the dependencies:

   pip install -r requirements.txt

3. Install the package:

   python setup.py install

Testing installation with an example

After you finish installing, you can test REVEALER by running

REVEALER test 

This will take approximately an hour.

Jupyter notebook Usage

Detailed example of using eREVEALER in Jupyter Notebook can be found here. eREVEALER is also available in GenePattern, allowing you to run it directly on the GenePattern server. More details can be found [here](link to genepattern module to be added).

Command line Usage

The preprocessing step offers various modes, which are explained in detail in the GenePattern documentation. Below are example commands for different modes.

Here is the command-line version of the example found here.

Download Example Input File

First, download the example input file for the CCLE dataset MAF file from this link: DepMap Public 23Q2 OmicsSomaticMutations.csv. Save it to the example/sample_input folder (or another location, as long as you indicate the path in the command).

Run File Preprocessing

REVEALER preprocess \
    --mode class \
    --input_file example/sample_input/OmicsSomaticMutations.csv \
    --protein_change_identifier ProteinChange \
    --file_separator , \
    --col_genename HugoSymbol \
    --col_class VariantType \
    --col_sample ModelID \
    --prefix CCLE \
    --out_folder example/sample_input/CCLE \
    --mode mutall

Convert Annotation from DepMap to CCLE

python example/DepMapToCCLE.py example/sample_input/NameConvert.csv example/sample_input/CCLE_Mut_All.gct example/sample_input/CCLE_Mut_All_rename.gct

Run REVEALER with Generated File and NFE2L2 Signature

REVEALER run \
    --target_file example_notebook/sample_input/CCLE_complete_sigs.gct \
    --feature_file example_notebook/sample_input/CCLE_Mut_All_rename.gct \
    --out_folder example_notebook/sample_output/NRF2 \
    --prefix CCLE_NRF2 \
    --target_name NFE2L2.V2 \
    --if_pvalue False \
    --if_bootstrap False \
    --gene_locus example_notebook/sample_input/allgeneLocus.txt \
    --tissue_file example_notebook/sample_input/TissueType_CCLE.gct

Contributing

If you would like to contribute to eREVEALER, please submit a pull request or report issues on our GitHub repository.

License

eREVEALER is licensed under the MIT License. See the LICENSE file for more details.