/chromWaveR

R package to predict nucleosome(/TF) occupancy from DNA sequence using pre-trained chromWave models.

Primary LanguageR

chromWaveR: R package to predict nucleosome occupancy from DNA sequence using pre-trained chromWave models.

Abstract

Requirements

  • Install following python packages:

    # CPU version 
pip install tensorflow==2 keras numpy
# GPU version 
pip install tensorflow-gpu==2 keras numpy

  • ALTERNATIVE: Setup Tensorflow/Keras CONDA env:

    1. Download and install CONDA.

    2. Install python packages

    conda create -n keras python
# CPU version 
pip install tensorflow==2 keras numpy
# GPU version 
pip install tensorflow-gpu==2 keras numpy

  • Install following R packages from CRAN:

    install.packages(c('tidyverse', 'reticulate', 'kerasR'))

  • Install following R packages from Bioconductor:

    install.packages(c('tidyverse', 'reticulate', 'kerasR'))

  • Install chromWaveR:

    if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("Biostrings")

Tutorial

For an example of how to use the package see here.

Reference

ChromWave: Deciphering the DNA-encoded competition between DNA-binding factors with deep neural networks

Preprint: Add link upon submission.