Repository associated with the manuscript “Boosting Graph Neural Networks with Molecular Mechanics: A Case Study of Sigma Profile Prediction”.

Repository structure:

• Main
  • Databases: graph and property database files.
  • HyperparameterSearch: hyperparameter tuning results of each GCN model.
  • Models: GCN models, train/val/test splits of final fitting, and CNN-related files.
  • Python: main Python code used throughout the manuscript. Files are numbered in chronological order. Details and instructions included.