Research map / Probabilistic and scientific modeling / Molecules and proteins
DCFold: Efficient Protein Structure Generation with Single Forward Pass
Zhe Zhang, Yuanning Feng, Yuxuan Song, Keyue Qiu, Hao Zhou, Wei-Ying Ma
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 |
|---|---|---|---|
| apply Temporal Geodesic Matching instead of ECM schedule (ablation) | consistency distillation of AF3 · PoseBusters V2 | success rate (%): 75.7 -> 77.5 | Improved |
| use naive consistency distillation (CD) schedule (ablation) | AF3 diffusion module · PoseBusters V2 | success rate (%): 25.6 | Hurt |
| apply dual consistency distillation (DCFold) to one-step AF3 ODE | Protenix AF3 · PoseBusters V2 | Best RMSD<1A (%): 51.43 -> 58.10 | Improved |
| apply dual consistency distillation to one-step AF3 ODE | Protenix AF3 · PoseBusters V2 | Worst RMSD<1A (%): 37.62 -> 46.67 | Improved |
| apply dual consistency distillation vs full AlphaFold3 | DCFold · PoseBusters V2 | Best RMSD<1A (%): 67.14 -> 58.10 | Hurt |
| distill to DCFold vs AF3 ODE | Protenix AF3 · Recent PDB PP-complex | success rate (%): 87.0 -> 92.2 | Improved |
| distill to DCFold vs AF3 ODE | Protenix AF3 · Recent PDB monomer | TM-score: 0.830 -> 0.850 | Improved |
| distill to DCFold vs full AF3 | Protenix AF3 · Recent PDB PL-complex | TM-score: 0.810 -> 0.824 | Improved |
| use DCFold instead of AF2 in BindCraft hallucination | BindCraft · Binder design average | physics-based success rate: .26 -> .29 | Improved |
| use DCFold instead of AF2 in BindCraft hallucination | BindCraft · Binder design average | model-based success rate: .69 -> .78 | Improved |
| use DCFold instead of AF2 in BindCraft | BindCraft · ALK physics-based | success rate: .14 -> .12 | Hurt |
| distill AF3 to DCFold | AlphaFold3 · PoseBusters V2 diversity | pairwise TM-score (lower better): 0.9646 -> 0.9701 | No clear effect |
Closest papers by meaning
- Energy-Based Flow Matching for Generating 3D Molecular Structure (ICML 2025 · similarity 0.89)
- Few-Step Cofolding with All-Atom Flow Maps (NeurIPS 2026 · similarity 0.88)
- SE(3)-Stochastic Flow Matching for Protein Backbone Generation (ICLR 2024 · similarity 0.87)
- AlphaFold Meets Flow Matching for Generating Protein Ensembles (ICML 2024 · similarity 0.87)
- Proteina: Scaling Flow-based Protein Structure Generative Models (ICLR 2025 · similarity 0.87)
- P(all-atom) Is Unlocking New Path For Protein Design (ICML 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.
