Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/rmeli/gnina-torch
🔥 PyTorch implementation of GNINA scoring function for molecular docking
https://github.com/rmeli/gnina-torch
deep-learning docking drug-discovery gnina pytorch pytorch-ignite scoring-functions
Last synced: 22 days ago
JSON representation
🔥 PyTorch implementation of GNINA scoring function for molecular docking
- Host: GitHub
- URL: https://github.com/rmeli/gnina-torch
- Owner: RMeli
- License: mit
- Created: 2021-10-19T16:28:50.000Z (about 3 years ago)
- Default Branch: main
- Last Pushed: 2023-05-16T06:29:15.000Z (over 1 year ago)
- Last Synced: 2024-01-27T17:44:35.241Z (9 months ago)
- Topics: deep-learning, docking, drug-discovery, gnina, pytorch, pytorch-ignite, scoring-functions
- Language: Python
- Homepage: https://gnina-torch.rtfd.io
- Size: 32.6 MB
- Stars: 45
- Watchers: 3
- Forks: 3
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
- Contributing: .github/CONTRIBUTING.md
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md
- Citation: CITATION.cff
Awesome Lists containing this project
README
# gnina-torch
[//]: # (Badges)
[](https://github.com/RMeli/gnina-torch/actions?query=workflow%3ACI)
[](https://codecov.io/gh/RMeli/gnina-torch)
[](https://gnina-torch.readthedocs.io/en/latest/?badge=latest)
[](https://zenodo.org/badge/latestdoi/419005477)
PyTorch implementation of [GNINA](https://github.com/gnina/gnina) scoring function.
> [!WARNING]
> [GNINA version `1.3`](https://github.com/gnina/gnina/releases/tag/v1.3) changed the deep learning backend from Caffe to PyTorch. Therefore, PyTorch models are now nativaly supported by [GNINA](https://github.com/gnina/gnina).
> Using [GNINA](https://github.com/gnina/gnina/releases/tag/v1.3) has the advantage that the models can be used directly within the docking pipeline, instead of being used for post-processing.
> The `gnina-torch` project is no longer under active development.
## References
```
@software{
gninatorch_2022,
author = {Meli, Rocco and McNutt, Andrew},
doi = {10.5281/zenodo.6943066},
month = {7},
title = {{gninatorch}},
url = {https://github.com/RMeli/gnina-torch},
version = {0.0.1},
year = {2022}
}
```
If you are using `gnina-torch`, please consider citing the following references:
> Protein-Ligand Scoring with Convolutional Neural Networks,
> M. Ragoza, J. Hochuli, E. Idrobo, J. Sunseri, and D. R. Koes, *J. Chem. Inf. Model.* 2017, 57 (4), 942-957.
> DOI: [10.1021/acs.jcim.6b00740](https://pubs.acs.org/doi/full/10.1021/acs.jcim.6b00740)
> libmolgrid: Graphics Processing Unit Accelerated Molecular Gridding for Deep Learning Applications
> J. Sunseri and D. R. Koes, *J. Chem. Inf. Model.* 2020, 60 (3), 1079-1084.
> DOI: [10.1021/acs.jcim.9b01145](https://pubs.acs.org/doi/10.1021/acs.jcim.9b01145)
If you are using the pre-trained `default2018` and `dense` models from [GNINA](https://github.com/gnina/gnina), please consider citing the following reference as well:
> Three-Dimensional Convolutional Neural Networks and a Cross-Docked Data Set for Structure-Based Drug Design,
> P. G. Francoeur, T. Masuda, J. Sunseri, A. Jia, R. B. Iovanisci, I. Snyder, and D. R. Koes, *J. Chem. Inf. Model.* 2020, 60 (9), 4200-4215.
> DOI: [10.1021/acs.jcim.0c00411](https://doi.org/10.1021/acs.jcim.0c0041)
If you are using the pre-trained `default` model ensemble from [GNINA](https://github.com/gnina/gnina), please consider citing the following reference as well:
> GNINA 1.0: molecular docking with deep learning,
> A. T. McNutt, P. Francoeur, R. Aggarwal, T. Masuda, R. Meli, M. Ragoza, J. Sunseri, D. R. Koes,
> *J. Cheminform.* 2021, 13 (43).
> DOI: [10.1186/s13321-021-00522-2](https://doi.org/10.1186/s13321-021-00522-2)
## Installation
The `gninatorch` Python package has several dependencies, including:
* [PyTorch](https://pytorch.org/)
* [PyTorch-Ignite](https://pytorch.org/ignite/)
* [libmolgrid](https://gnina.github.io/libmolgrid/)
A full developement environment can be installed using the [conda](https://docs.conda.io/en/latest/) package manager and the provided [conda](https://docs.conda.io/en/latest/) environment file (`devtools/conda-envs/gninatorch.yaml`):
```bash
conda env create -f devtools/conda-envs/gninatorch.yaml
conda activate gninatorch
```
Once the [conda](https://docs.conda.io/en/latest/) environment is created and activated, the `gninatorch` package can be installed using [pip](https://pip.pypa.io/en/stable/) as follows:
```bash
python -m pip install .
```
### Tests
In order to check the installation, unit tests are provided and can be run with [pytest](https://docs.pytest.org/):
```bash
pytest --cov=gninatorch
```
## Usage
Training and inference modules try to follow the original [Caffe](https://caffe.berkeleyvision.org/) implementation of [gnina/scripts](https://github.com/gnina/scripts), however not all features are implemented.
The folder `examples` includes some complete examples for training and inference.
The folder `gninatorch/weights` contains pre-trained models from [GNINA](https://github.com/gnina/gnina), converted from Caffe to PyTorch.
### Pre-trained GNINA models
Pre-trained GNINA models can be loaded as follows:
```python
from gninatorch.gnina import setup_gnina_model
model = setup_gnina_model(MODEL)
```
where `MODEL` corresponds to the `--cnn` argument in [GNINA](https://github.com/gnina/gnina).
A single model will return `log_CNNscore` and `CNNaffinity`, while an ensemble of models will return `log_CNNscore`, `CNNaffinity`, and `CNNvariance`.
Inference with pre-trained [GNINA](https://github.com/gnina/gnina) models (`--cnn` argument in [GNINA](https://github.com/gnina/gnina)) is implemented in the `gnina` module:
```bash
python -m gninatorch.gnina --help
```
### Training
Training is implemented in the `training` module:
```bash
python -m gninatorch.training --help
```
### Inference
Inference is implemented in the `inference` module:
```bash
python -m gninatorch.inference --help
```
## Acknowledgments
Project based on the [Computational Molecular Science Python Cookiecutter](https://github.com/molssi/cookiecutter-cms) version 1.6.
The pre-trained weights of [GNINA](https://github.com/RMeli/gnina-torch) converted to PyTorch were kindly provided by Andrew McNutt (@drewnutt).
---
Copyright (c) 2021-2022, Rocco Meli