Research map / Probabilistic and scientific modeling / Molecules and proteins
Learning Gradient Fields for Molecular Conformation Generation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Generate conformations directly in coordinate space rather than generate distances and post-process (ablation) | ConfGF · GEOM-QM9 | Mean COV (%): 81.94 -> 88.49 | Helped |
| Generate conformations directly in coordinate space rather than generate distances and post-process (ablation) | ConfGF · GEOM-QM9 | Mean MAT (Å): 0.3861 -> 0.2673 | Helped |
| Generate conformations directly in coordinate space rather than generate distances and post-process (ablation) | ConfGF · GEOM-Drugs | Mean COV (%): 53.40 -> 62.15 | Helped |
| Generate conformations directly in coordinate space rather than generate distances and post-process (ablation) | ConfGF · GEOM-Drugs | Mean MAT (Å): 1.2493 -> 1.1629 | Helped |
| Use ConfGF rather than RDKit for conformation generation | Molecular conformation generation · GEOM-QM9 | Mean COV (%): 83.26 -> 88.49 | Helped |
| Use ConfGF rather than RDKit for conformation generation | Molecular conformation generation · GEOM-Drugs | Mean COV (%): 60.91 -> 62.15 | Helped |
| Use ConfGF rather than CGCF for single-distance distribution modeling | ConfGF · ISO17, Single distances | Median MMD: 0.1786 -> 0.2358 | Hurt |
| Use ConfGF rather than CGCF for pair-distance distribution modeling | ConfGF · ISO17, Pair distances | Median MMD: 0.2734 -> 0.3206 | Hurt |
| Use ConfGF rather than RDKit to predict average ensemble energy | Ensemble property prediction · GEOM-QM9 | MAE average energy (eV): 0.9233 -> 2.7886 | Hurt |
| Use ConfGF rather than RDKit to predict minimum ensemble energy | Ensemble property prediction · GEOM-QM9 | MAE minimum energy (eV): 0.6585 -> 0.1765 | Helped |
| Use ConfGF rather than RDKit to predict average HOMO-LUMO gap | Ensemble property prediction · GEOM-QM9 | MAE average HOMO-LUMO gap (eV): 0.3698 -> 0.4688 | Hurt |
| Use ConfGF rather than RDKit to predict maximum HOMO-LUMO gap | Ensemble property prediction · GEOM-QM9 | MAE maximum HOMO-LUMO gap (eV): 0.2359 -> 0.1433 | Helped |
Closest papers by meaning
- Swallowing the Bitter Pill: Simplified Scalable Conformer Generation (ICML 2024 · similarity 0.92)
- DiSCO: Diffusion Schrödinger Bridge for Molecular Conformer Optimization (AAAI 2024 · similarity 0.88)
- Score-based 3D molecule generation with neural fields (NeurIPS 2024 · similarity 0.88)
- WGFormer: An SE(3)-Transformer Driven by Wasserstein Gradient Flows for Molecular Ground-State Conformation Prediction (ICML 2025 · similarity 0.88)
- Enhancing Generalizability in Molecular Conformation Generation with METRIZATION-Informed Geometric Diffusion Pretraining (AAAI 2025 · similarity 0.88)
- REBIND: Enhancing Ground-state Molecular Conformation Prediction via Force-Based Graph Rewiring (ICLR 2025 · similarity 0.87)
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.