Research map / Graph learning / GNN architectures
MMPG: MoE-based Adaptive Multi-Perspective Graph Fusion for Protein Representation Learning
Yusong Wang, Jialun Shen, Zhihao Wu, Yicheng Xu, Shiyin Tan, Mingkun Xu, Changshuo Wang, Zixing Song, Prayag Tiwari
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 geometric-structural perspective (ablation) | MMPG · EC | Fmax: 0.821 -> 0.893 | Improved |
| add chemical-functional perspective (ablation) | MMPG · EC | Fmax: 0.851 -> 0.893 | Improved |
| add physical-energetic perspective (ablation) | MMPG · EC | Fmax: 0.874 -> 0.893 | Improved |
| add physical-energetic perspective (ablation) | MMPG · Enzyme Reaction | accuracy: 84.9 -> 89.0 | Improved |
| use MoE fusion instead of edge stacking (ablation) | MMPG · EC | Fmax: 0.794 -> 0.893 | Improved |
| use MoE fusion instead of concatenation (ablation) | MMPG · GO-MF | Fmax: 0.600 -> 0.663 | Improved |
| add auxiliary classification loss (ablation) | MMPG · FOLD Fold | accuracy: 58.8 -> 60.9 | Improved |
| multi-perspective graph vs single radius graph | MMPG · EC | Fmax: 0.840 -> 0.893 | Improved |
| MMPG vs ESM-2 and EPGGCL on FOLD Superfamily | MMPG · FOLD Superfamily | accuracy: 80.8 (EPGGCL) vs 79.5 | Hurt |
| increase number of experts beyond 10 (ablation) | MMPG · GO, FOLD, EC | performance | Hurt |
| mask 40% of residue embeddings (ablation) | MMPG · EC, FOLD | performance | Hurt |
Closest papers by meaning
- Boosting Protein Graph Representations through Static-Dynamic Fusion (ICML 2025 · similarity 0.84)
- Enhancing Graph Contrastive Learning for Protein Graphs from Perspective of Invariance (ICML 2025 · similarity 0.84)
- Towards Multiscale Graph-based Protein Learning with Geometric Secondary Structural Motifs (NeurIPS 2025 · similarity 0.84)
- PRING: Rethinking Protein-Protein Interaction Prediction from Pairs to Graphs (NeurIPS 2025 · similarity 0.83)
- M²N: A Progressive Macro-to-Micro 3D Modeling Scheme for Unveiling Drug-Target Affinity (AAAI 2025 · similarity 0.83)
- Evaluating Representation Learning on the Protein Structure Universe (ICLR 2024 · similarity 0.83)
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.
