AFSE

Reproducing of the paper entitled "AFSE: towards improving model generalization of deep graph learning of ligand bioactivities targeting GPCR proteins" (Briefings in Bioinformatics, 2022, IF: 13.994)

• All rights reserved by Yueming Yin, Email: [email protected] (or [email protected]).
• AFSE has been deployed on our web page: www.noveldelta.com/AFSE.

The Code Ocean compute capsule will allow you to reproduce the results published by the author on your local machine¹. Follow the instructions below, or consult the knowledge base for more information. Don't hesitate to reach out via live chat or email if you have any questions.

¹ You may need access to additional hardware and/or software licenses.

Prerequisites

• Docker Community Edition (CE)
• nvidia-docker for code that leverages the GPU
• Licenses where applicable

Instructions

The computational environment (Docker image)

This capsule is private and its environment cannot be downloaded at this time. You will need to rebuild the environment locally.

If there's any software requiring a license that needs to be run during the build stage, you'll need to make your license available. See the knowledge base for more information.

In your terminal, navigate to the folder where you've extracted the capsule and execute the following command:

cd environment && docker build . --tag AFSE; cd ..

This step will recreate the environment (i.e., the Docker image) locally, fetching and installing any required dependencies in the process. If any external resources have become unavailable for any reason, the environment will fail to build.

Running the capsule to reproduce the results

In your terminal, navigate to the folder where you've extracted the capsule and execute the following command, adjusting parameters as needed:

nvidia-docker run --it \
  --workdir /AFSE \
  --volume "$PWD/data":/Benchmark_Datasets \
  --volume "$PWD/code":/AFSE \
  AFSE

Reproduction on Attentive FP

Reproducing the training process of AFSE

In your jupyter notebook, set the task ID to reproduce the training process of AFSE using the data in "Benchmark_Datasets":

./AFSE/Train_AFSE.ipynb

Reproducing the test and visualization process of AFSE

In your jupyter notebook, set the task ID to reproduce the training process of AFSE using the data in "Benchmark_Datasets" and the final models in "AFSE_Models":

./AFSE/Test&Viz_AFSE.ipynb

Reproduction on other models

Please refer to "Molecular Property Prediction on a New CSV Dataset"

Citations

In Latex

@article{yin2022afse,
  title={AFSE: towards improving model generalization of deep graph learning of ligand bioactivities targeting GPCR proteins},
  author={Yin, Yueming and Hu, Haifeng and Yang, Zhen and Jiang, Feihu and Huang, Yihe and Wu, Jiansheng},
  journal={Briefings in Bioinformatics},
  volume={23},
  number={3},
  pages={bbac077},
  year={2022},
  publisher={Oxford University Press}
}

In Word

Yin, Yueming, Haifeng Hu, Zhen Yang, Feihu Jiang, Yihe Huang, and Jiansheng Wu. "AFSE: towards improving model generalization of deep graph learning of ligand bioactivities targeting GPCR proteins." Briefings in Bioinformatics 23, no. 3 (2022): bbac077.