Multi_CycGT: A DL-Based Multimodal Model for Membrane Permeability Prediction of Cyclic Peptides

This is the code for "Multi_CycGT: A DL-Based Multimodal Model for Membrane Permeability Prediction of Cyclic Peptides" paper.

Directory Structure

bashCopy code/
├── data/            
├── data_process/  
├── model/         
├── LICENSE        
└── README.md       

Quick-start

Install dependency

Running this command for installing dependency in docker:

pip install requirments.txt
./replace.sh

Directory structure

Before start training the model, you need to process the dataset. The /data contains datasets and preprocessing code.

Training model

Running this command for training the Multi_CycGT model：

python ./model/deep_learing/model_concat.py

Membrane permeability prediction

Running this command for predicting membrane permeability of cyclic peptides:

python ./model/deep_learing/model_concat.py

Example

we provide a detailed example in the notebook： Notebook.

Support or Report Issues

If you encounter any issues or need support while using Multi_CycGT, please report the issue in the GitHub Issues .

Copyright and License

The project of Multi_CycGT follows MIT License. Please read the license carefully before use.

Update History

• v1.0.0 (2023-11-20):

  The first official version is released.

  • Supporting multi-mode data input.

  • Realizing the membrane permeability prediction.

  • Adding an example notebook.

Related Projects

The related projects about membrane permeability prediction and deep learning.

• Related Project 1
• Related Project 2

Model Scalability

Running time of Multi_CycGT