/epridict

Chromatin context dependent prediction of prime editing rates (ePRIDICT)

Primary LanguagePythonMIT LicenseMIT

ePRIDICT: epigenetic PRIme editing preDICTion

Overview

This repository is part of our study in Nature Biotechnology "Machine learning prediction of prime editing efficiency across diverse chromatin contexts" and the initial BioRxiv preprint.

Predict prime editing efficiency based on chromatin context of a genomic location in K562 cells. Repository containing Python package for running trained ePRIDICT (epigenetic PRIme editing preDICTion) models. Models were trained in K562 cells. Prediction performance may vary in different cellular contexts. Check out our publication for further details.

Complementary Model

  • PRIDICT2.0: This model focuses on the sequence-context based prime efficiency prediction and pegRNA design. We recommend to first select the most suitable pegRNA with PRIDICT2.0 and then assess its overall endogenous targetability with ePRIDICT. Access PRIDICT2.0 GitHub Repository

Resources

  • Supplementary Files: Access Here
  • Web Application: For an online version of ePRIDICT, visit our webapp*.

*Default model for this repository and online webapp is ePRIDICT-light. For running the full ePRIDICT model, check the description below.

Contact

For questions or suggestions, please either:

Citation

If find our work useful for your research please cite:

Getting Started

Installation using Anaconda (Linux and Mac OS) 🐍

📣 ePRIDICT can only be installed on Linux and Mac OS since pybigwig package is not available for Windows 📣

The easiest way to install and manage Python packages on various OS platforms is through Anaconda. Once installed, any package (even if not available on Anaconda channel) could be installed using pip.

  • Install Anaconda.

  • Start a terminal and run:

    # clone ePRIDICT repository
git clone https://github.com/Schwank-Lab/epridict.git
# navigate into repository
cd epridict
# create conda environment and install dependencies for ePRIDICT (only has to be done before first run/install)
# use epridict_linux for linux machine or epridict_mac for a macbook
conda env create -f epridict_linux.yml # epridict_mac.yml for macbook

# activate the created environment
conda activate epridict

  • Next, downloading ENCODE datasets is needed for prediction with ePRIDICT. Files will be downloaded in bigwig folder. Note: For running the full model, 455 datasets will be downloaded, requiring 624 GB of storage space! For the light model, with near on-par performance, 6 datasets will be downloaded, requiring 5.3 GB of storage space.

    # make download script executable:
chmod +x epridict_download_encode.sh
# run download script:
./epridict_download_encode.sh light # or ./epridict_download_encode.sh full

Running ePRIDICT in 'manual' mode:

Required:

  • --chromosome: Chromosome of desired location. Format: "chr1", "chr2", ...
  • --position_hg38: Position within chromosome (hg38). Example: "1192940"

Optional:

  • --use_full_model: Use full model (455 ENCODE datasets) for prediction. Only possible when downloaded all datasets with ./epridict_download_encode.sh full. Default: light model
python epridict_prediction.py manual --chromosome chr3 --position_hg38 44843504
# for full model:
# python epridict_prediction.py manual --chromosome chr3 --position_hg38 44843504 --use_full_model

Running in 'batch' mode:

Required:

  • input_filename: Input file name - name of .csv file that has two columns [chromosome, position_hg38]. See sample_epridict_batch.csv in the ./input folder.

Optional:

  • --output-fname: Alternative output filename. Default is input_filename_output.csv.
  • --use_full_model: Use full model (455 ENCODE datasets) for prediction. Only possible when downloaded all datasets with ./epridict_download_encode.sh full. Default: light model
python epridict_prediction.py batch sample_epridict_batch.csv
# for full model and alternative output filename:
# python epridict_prediction.py batch sample_epridict_batch.csv --output-fname alternative_output_filename_batch.csv --use_full_model