Research map / Vision-language and multimodal / Human and embodied modeling
Learning Cell-Aware Hierarchical Multi-Modal Representations for Robust Molecular Modeling
Mengran Li, Zelin Zang, Wenbin Xing, Junzhou Chen, Ronghui Zhang, Jiebo Luo, Stan Z. Li
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 structure-aware propagation instead of zero imputation (ablation) | CHMR · ChEMBL | AUC (%): 81.6 -> 84.7 | Improved |
| add structure-aware propagation instead of random imputation (ablation) | CHMR · Biogen | MAE x100: 44.1 -> 40.9 | Improved |
| use graph propagation instead of neighbor imputation (ablation) | CHMR · ChEMBL | AUC (%): 83.1 -> 84.7 | Improved |
| add semantic consistency alignment (SCA) (ablation) | CHMR · ChEMBL | AUC (%): 82.4 -> 84.7 | Improved |
| add distribution-level alignment (ablation) | CHMR · Biogen | MAE x100: 41.9 -> 40.9 | Improved |
| add sample-level alignment (ablation) | CHMR · Biogen | MAE x100: 42.6 -> 40.9 | Improved |
| add tree-structured VQ (ablation) | CHMR · ChEMBL | AUC (%): 82.3 -> 84.7 | Improved |
| replace flat VQ with tree-structured VQ (ablation) | CHMR · Biogen | MAE x100: 42.0 -> 40.9 | Improved |
| add context-propagation reconstruction (CPR) (ablation) | CHMR · ChEMBL | AUC (%): 82.6 -> 84.7 | Improved |
| add random walks in biological graph (ablation) | CHMR · Biogen | MAE x100: 42.1 -> 40.9 | Improved |
| add cellular modalities to molecular features (ablation) | CHMR · Biogen | MAE x100: 44.3 -> 40.9 | Improved |
| use only gene or cell modality with molecules instead of all three (ablation) | CHMR · ChEMBL | AUC (%): 82.7 -> 84.7 | Improved |
Closest papers by meaning
- Learning Molecular Representation in a Cell (ICLR 2025 · similarity 0.88)
- Atomas: Hierarchical Adaptive Alignment on Molecule-Text for Unified Molecule Understanding and Generation (ICLR 2025 · similarity 0.87)
- How Molecules Impact Cells: Unlocking Contrastive PhenoMolecular Retrieval (NeurIPS 2024 · similarity 0.87)
- KnowMol: Advancing Molecular Large Language Models with Multi-Level Chemical Knowledge (NeurIPS 2025 · similarity 0.86)
- Omni-Mol: Multitask Molecular Model for Any-to-any Modalities (NeurIPS 2025 · similarity 0.86)
- TRIDENT: Tri-Modal Molecular Representation Learning with Taxonomic Annotations and Local Correspondence (NeurIPS 2025 · 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.
