ProBID-NET is a deep-learning model for designing amino acid on protein-protein interfaces.

Installation (tested on Linux)

1. download the source code

run git clone https://github.com/ComputArtCMCG/ProBID-NET.git to clone the repository.
Alternatively, download the zip file, unzip it and and navigate to the folder using the following commands:
unzip ProBID-NET-main.zip && cd ./ProBID-NET-main/

2. download the checkpoint of the model

The checkpoint of model trained on both Chain-chain interface and domain-domain interface sets is available at https://figshare.com/s/ebbd5184c0a46fb2b179, download modeloutput0.hdf5 and move it to the model directory in ProBID-NET-main.

3. prepare anaconda environment

Install anaconda from https://anaconda.org/ if it is not installed in the system.
Run the following commands to setup the env:
conda create -n keras2.8.0 python=3.9
conda activate keras2.8.0
conda install keras=2.8.0
pip install pandas biopython

4. make prediction

Run bash ./ProBID-Net_run.sh examples/1euv.input ./examples/ to make prediction on an test protein.
The prediction output is saved in examples/output_pred.
The first option of ProBID-Net_run.sh is a file containing list of PDB file and chain ID to predict.
The second option points to a directory where PDB files are saved.

Dataset

The lists of protein-protein complex structures for the training set and test sets are available in the dataset folder.

Reference

Zhihang Chen et. al, ProBID-Net: A Deep Learning Model for Protein-Protein Binding Interface Design, submitted

If you encounter any issues during installation, please open an issue.