BreakNet

A deep learning method to detect deletion from long reads alignment. It is built with Tensorflow and Python 3.

Installation

Requirements

python 3.7, numpy, scipy, pandas, Matplotlib, TensorFlow 2.4, pysam

1. Create a virtual environment

# create
conda create -n BreakNet python=3.6
# activate
conda activate BreakNet
# deactivate
conda deactivate

2. clone BreakNet

After creating and activating the BreakNet virtual environment, download BreakNet from github:
```
git clone https://github.com/luojunwei/BreakNet.git
cd BreakNet
```
3. Install

conda activate BreakNet
conda install numpy, scipy, pandas, Matplotlib, TensorFlow, pysam

Tested data

The example data can be downloaded from

Usage

Train Model

1. Produce data for training

python breaknet.py data_mode bamfile_path output_data_folder vcf_path

  bamfile_path is the path of the alignment file about the reference and the long read set. And, the bam file should be sorted and indexed;

  output_data_folder is a folder which is used to store training data or evaluation data;

  vcf_path is the path of the vcf which is used to label training data;

2. Train a new model

python breaknet.py train_mode training_data_folder evaluation_data_folder trained_weight_path epochs

  First, we use commond 1 to produce training data, which is stored in the training_data_folder.

  Second, we use commond 1 to produce evaluation data, which is stored in the evaluation_data_folder.

  trained_weight_path is the path of the trained weight file of the model. Not include extension name.

  epochs are max training epochs.

Call deletion

In the folder "trained_weight", we give the trained weight files, which can directly used for calling deletions.

1. Produce data for call sv

python breaknet.py data_mode bamfile_path call_folder

2. Call deletion

python breaknet.py call_mode call_folder trained_weight_path bamfilepath

luojunwei / BreakNet

readme