https://github.com/zetayue/MXMNet

Source code for "Molecular Mechanics-Driven Graph Neural Network with Multiplex Graph for Molecular Structures"
https://github.com/zetayue/MXMNet

computational-biology deep-learning drug-discovery geometric-deep-learning graph-neural-networks machine-learning

Last synced: 24 days ago
JSON representation

Source code for "Molecular Mechanics-Driven Graph Neural Network with Multiplex Graph for Molecular Structures"

• Host: GitHub
• URL: https://github.com/zetayue/MXMNet
• Owner: zetayue
• License: mit
• Created: 2020-11-20T19:15:31.000Z (about 4 years ago)
• Default Branch: master
• Last Pushed: 2023-11-15T02:53:59.000Z (about 1 year ago)
• Last Synced: 2024-08-09T13:20:35.642Z (4 months ago)
• Topics: computational-biology, deep-learning, drug-discovery, geometric-deep-learning, graph-neural-networks, machine-learning
• Language: Python
• Homepage:
• Size: 48.2 MB
• Stars: 70
• Watchers: 2
• Forks: 16
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• StarryDivineSky - zetayue/MXMNet

README

        
# Molecular Mechanics-Driven Graph Neural Network with Multiplex Graph for Molecular Structures

Code for **MXMNet** proposed in our paper: **[Molecular Mechanics-Driven Graph Neural Network with Multiplex Graph for Molecular Structures](https://arxiv.org/abs/2011.07457)**, which has been accepted by the *Machine Learning for Structural Biology Workshop* ([MLSB 2020](https://www.mlsb.io/)) and the *Machine Learning for Molecules Workshop* ([ML4Molecules 2020](https://ml4molecules.github.io/)) at the *34th Conference on Neural Information Processing Systems* (NeurIPS 2020).

### Important Update about Improved Model (2023/11)

We have released the **[code](https://github.com/XieResearchGroup/Physics-aware-Multiplex-GNN)** for **PAMNet** in our *Nature Scientific Reports* paper "**[A universal framework for accurate and efficient geometric deep learning of molecular systems](https://www.nature.com/articles/s41598-023-46382-8)**", which is an improved version of MXMNet with **higher accuracy and efficiency**. **We highly recommend anyone interested in MXMNet try our PAMNet**.

## Overall Architecture







## Requirements

CUDA : 10.1

Python : 3.7.10

The other dependencies can be installed with:

```

pip install -r requirements.txt

```

## How to Run

You can directly download, preprocess the QM9 dataset and train the model with 

```

python main.py

```

Optional arguments:

```

  --gpu             GPU number

  --seed            random seed

  --epochs          number of epochs to train

  --lr              initial learning rate

  --wd              weight decay value

  --n_layer         number of hidden layers

  --dim             size of input hidden units

  --batch_size      batch size

  --target          index of target (0~11) for prediction on QM9

  --cutoff          distance cutoff used in the global layer

```

The default model to be trained is the MXMNet (BS=128, d_g=5) by using '--batch_size=128 --cutoff=5.0'.

## Cite

If you find this model and code are useful in your work, please cite our paper:

```

@article{zhang2020molecular,

  title={Molecular mechanics-driven graph neural network with multiplex graph for molecular structures},

  author={Zhang, Shuo and Liu, Yang and Xie, Lei},

  journal={arXiv preprint arXiv:2011.07457},

  year={2020}

}

```