Datasets

---

Datasets used for training and testing

T145.zip : A dataset containing 145 proteins used for testing the prediction results of DeepGlycanSite.

T145-AF2.zip : A dataset containing AlphaFold-predicted structures of the 145 proteins used for testing the prediction results of DeepGlycanSite.

train_data.csv : The protein data source used for training DeepGlycanSite.

Codes of DeepGlycansite

---

Download from Github : Source for downloading the DeepGlycanSite code.

Download Files

• single_case_prediction.py : The main prediction script for running DeepGlycanSite predictions.
• DeepGlycanSite_lig.py : Functions required for predictions using the DeepGlycanSite+Ligand model.
• DeepGlycanSite.py : Functions required for predictions using the DeepGlycanSite model.
• environment.yml : Configuration file specifying package versions for setting up the environment.
• README.md : The README file providing an overview and instructions.
• model : A folder containing the internal logic and structure of the model.
• P2Y14_example : A test case folder including ligand and receptor data for P2Y14, along with test hyperparameters.
• unicore : The core component of UniMol, used for extracting ligand information.
• src : Files related to glycan processing.
• visualizations : Contains overview images.
• datasets : Folder for dataset files.
• ckpts : Folder for model checkpoints.