A hypergraph neural network with reaction-center-aware negative sampling achieves higher F1 scores than GCN/GAT/HGNN baselines for reaction virtual screening on USPTO subsets.
Molecular hypergraph neural networks
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
ChemHGNN: A Hierarchical Hypergraph Neural Network for Reaction Virtual Screening and Discovery
A hypergraph neural network with reaction-center-aware negative sampling achieves higher F1 scores than GCN/GAT/HGNN baselines for reaction virtual screening on USPTO subsets.