MolHGCN
A Hypergraph Convolutional Neural Network for Molecular Properties Prediction using Functional Group
Install / Use
/learn @sailfish009/MolHGCNREADME
MolHGCN
A Hypergraph Convolutional Neural Network for Molecular Properties Prediction using Functional Group [paper]
Requirements
dgl 0.5.3
dgllife 0.2.6
networkx 2.5
torch 1.7.0
scikit-learn 0.22.1
RDKit
Dataset
Datasets can be found in: https://bit.ly/34wuwJo
Quickstart
molhgcn_classification.pyandmolhgcn_regression.pyare for the classification and regression tasks of MolHGCN respectively.ablation_classification_atom.pyandablation_regression_atom.pyare for the MolHGCN-AtomGC experiments in Section 4.2.1 respectively.ablation_classification_fg.pyandablation_regression_fg.pyare for the MolHGCN-FuncGC experiments in Section 4.2.1 respectively.
View on GitHub67/100
Security Score
Audited on Oct 6, 2025
No findings
