Research map / Graph learning / Graph self-supervised learning
I2Mole: Interaction-aware Invariant Molecular Learning For Generalizable Property Prediction
Wenjie Du, Jiahui Zhang, Xuqiang Li, Sihan Wang, Zhengyang Zhou, hongxin xiang, Jun Xia, Ye Wei, Yang Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add VQ environment codebook (ablation) | I2Mole · ZhangDDI | ACC: 74.52 -> 88.64 | Improved |
| add VQ environment codebook (ablation) | I2Mole · ZhangDDI | AUROC: 83.61 -> 95.12 | Improved |
| add intermolecular interaction (ablation) | I2Mole · ZhangDDI | ACC: 84.51 -> 88.64 | Improved |
| add intermolecular interaction (ablation) | I2Mole · ZhangDDI | AUROC: 87.21 -> 95.12 | Improved |
| add graph information bottleneck (ablation) | I2Mole · ZhangDDI | ACC: 84.72 -> 88.64 | Improved |
| add graph information bottleneck (ablation) | I2Mole · ZhangDDI | F1: 81.07 -> 85.87 | Improved |
| set beta (LMI weight) to 1E-4 instead of 0 (ablation) | I2Mole · ZhangDDI | ACC: 85.71 -> 88.64 | Improved |
| increase beta to 0.1 (ablation) | I2Mole · ZhangDDI | ACC: 88.64 -> 56.46 | Hurt |
| vary gamma (codebook loss weight) from 2E-5 to 1E-3 (ablation) | I2Mole · ZhangDDI | ACC: 88.31, 88.64, 88.21, 88.64, 88.32 | No clear effect |
| use I2Mole instead of IGIB-ISE | transductive DDI · DeepDDI | AUROC: 98.02 -> 99.04 | Improved |
| use I2Mole instead of IGIB-ISE in domain shift | trained on ZhangDDI · ChchMiner | ACC: 57.06 -> 59.25 | Improved |
| use I2Mole instead of IGIB-ISE on unseen-unseen drug pairs | inductive Type2 · ChchMiner | ACC: 68.92 vs 69.23 | No clear effect |
Closest papers by meaning
- Reaction Prediction via Interaction Modeling of Symmetric Difference Shingle Sets (NeurIPS 2025 · similarity 0.87)
- Learning Over Molecular Conformer Ensembles: Datasets and Benchmarks (ICLR 2024 · similarity 0.87)
- Association Pattern-enhanced Molecular Representation Learning (AAAI 2025 · similarity 0.87)
- Generalist Equivariant Transformer Towards 3D Molecular Interaction Learning (ICML 2024 · similarity 0.87)
- TRIDENT: Tri-Modal Molecular Representation Learning with Taxonomic Annotations and Local Correspondence (NeurIPS 2025 · similarity 0.86)
- CausalBind: Causal Modeling and Learning for Protein-Molecule Virtual Screening (NeurIPS 2026 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
