https://github.com/cxhernandez/molencoder

Molecular AutoEncoder in PyTorch
https://github.com/cxhernandez/molencoder

autoencoder computational-chemistry deep-learning python pytorch

Last synced: 20 days ago
JSON representation

Molecular AutoEncoder in PyTorch

• Host: GitHub
• URL: https://github.com/cxhernandez/molencoder
• Owner: cxhernandez
• License: mit
• Created: 2017-05-20T03:14:59.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2020-05-13T19:23:25.000Z (over 4 years ago)
• Last Synced: 2024-10-14T06:24:23.599Z (about 1 month ago)
• Topics: autoencoder, computational-chemistry, deep-learning, python, pytorch
• Language: Python
• Size: 49.8 MB
• Stars: 86
• Watchers: 6
• Forks: 17
• Open Issues: 7
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# MolEncoder

Molecular AutoEncoder in PyTorch

## Install

```bash

$ git clone https://github.com/cxhernandez/molencoder.git && cd molencoder

$ python setup.py install

```

## Download Dataset

```bash

$ molencoder download --dataset chembl22

```

## Train

```bash

$ molencoder train --dataset data/chembl22.h5

```

Add `--cuda` flag to enable CUDA. Add `--cont` to continue training a model from a checkpoint file.

## Pre-Trained Model

A pre-trained reference model is available in the `ref/` directory. Currently, it performs with ~98% accuracy on the validation set after 100 epochs of training. However, if you succeed at training a better model, feel free to submit a pull request!

## TODO

- [x] Implement encoder

- [x] Implement decoder

- [x] Add download command

- [x] Add train command

- [ ] Add encode command

- [ ] Add decode command

- [x] Add pre-trained model

## Shoutouts

+ [Original paper](https://arxiv.org/abs/1610.02415) by Gómez-Bombarelli, et al.

+ [keras-molecules](https://github.com/maxhodak/keras-molecules) by Max Hodak

+ [DeepChem](https://github.com/deepchem/deepchem)

+ [PyTorch](pytorch.org)