PAULYKnowledge Hub
Graph-Convolutional Neural Network (GNN)
Learns directly from a molecule's structural graph — powerful, but data-hungry.
Group-Contribution Approach (GCA)
Builds a prediction from known molecular fragments — transparent, but limited to familiar structures.
k-Nearest Neighbors (kNN)
Predicts from the most similar known molecules — robust, but only as good as its nearest neighbour.