schnet

EGNN_Network (lucidrains) being implemented with out-of-the-box (PyTorch) multihead attention on synthetic (PB), rigid-body protein pK-shift chemical labels. piKa predecessor, pKaSchNet realization (physics-informed, molecular ML: 121K pKPDB-RCSB PDBs, all-atom). No license, all rights reserved. Part of my FUB MSc. Compute credit: NHR/ZIB - A100.

Graph Neural Networks (SchNet & DMPNN) for HOMO-LUMO gap prediction of quinone derivatives — DFT B3LYP/6-31G* dataset, Optuna HPO, stratified 5-fold CV

ARGA,ARGVA,AttentiveFP,Captum,ComplEx,CorrectAndSmooth,DeepGCNLayer,DeepGraphInfomax,DimeNet,DimeNetPlusPlus,DistMult,EdgeCNN,GAE,GAT,GATv2,GCN,GITMol,GLEM,GIN,GNNFF,GraphSAGE,GraphUNet,GRetriever,HeteroJumpingKnowledge,InnerProductDecoder,JumpingKnowledge,KGEModel,LabelPropagation,LightGCN,LINKX,MetaLayer,MetaPath2Vec,MLP,MoleculeGPT,Node2Vec,

From-scratch SchNet-style message-passing neural network potential for BCC TiZrNb, benchmarked head to head against descriptor baselines at matched data and budget. Surrogate teacher labels (CHGNet), real-DFT anchors, leakage-safe splits, autograd forces, NVE stability.

Predicting Amino Acids' Dipole Moment Value using SchNet

SchNetPack - Deep Neural Networks for Atomistic Systems