Research map / Probabilistic and scientific modeling / Molecules and proteins
Co-Generative De Novo Functional Protein Design
Xinrui Chen, Yizhen Luo, Siqi Fan, Zaiqing Nie
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 |
|---|---|---|---|
| replace one-step sequence generation with co-generation of sequence and structure (without FSR/LSFS) (ablation) | CodeFP · GO-conditioned design | F1-Macro: 0.370 (CFP-Gen) -> 0.400 | Improved |
| use co-generation instead of two-step generation (ablation) | CodeFP · GO-conditioned design | pLDDT > 70 success rate (%): 52.24 -> 71.71 | Improved |
| add Functional-Structural Retrieval (FSR) (ablation) | CodeFP · GO-conditioned design | pLDDT > 70 success rate (%): 71.57 -> 80.65 | Improved |
| add FSR (ablation) | CodeFP · GO-conditioned design | F1-Macro: 0.423 -> 0.446 | Improved |
| add FSR on top of LSFS-less model (ablation) | CodeFP w/o LSFS · GO-conditioned design | F1-Micro: 0.465 -> 0.495 | Improved |
| add Local Structure-Function Supervision (LSFS) (ablation) | CodeFP · GO-conditioned design | MMD: 0.172 -> 0.106 | Improved |
| add LSFS (ablation) | CodeFP · GO-conditioned design | F1-Macro: 0.437 -> 0.446 | Improved |
| add LSFS (ablation) | CodeFP · GO-conditioned design | pLDDT > 70 success rate (%): 82.01 -> 80.65 | Hurt |
| add FSR (ablation) | CodeFP · GO-conditioned design | MRR: 0.674 -> 0.658 | Hurt |
| co-generation vs best baselines | CodeFP · Foldability (ESMFold) | pLDDT > 70 success rate (%): 74.22 -> 80.65 | Improved |
| CodeFP vs baseline on hypothetical combinations | CodeFP · Hypothetical functional combinations | Partial Match Rate (%): 5.54% -> 43.20% | Improved |
| CodeFP distributional alignment vs ProteoGAN | CodeFP · GO-conditioned design | MMD: 0.106 vs 0.095 | No clear effect |
Closest papers by meaning
- Protein Design with Dynamic Protein Vocabulary (NeurIPS 2025 · similarity 0.91)
- CFP-Gen: Combinatorial Functional Protein Generation via Diffusion Language Models (ICML 2025 · similarity 0.89)
- Sequence-Augmented SE(3)-Flow Matching For Conditional Protein Generation (NeurIPS 2024 · similarity 0.88)
- CarbonNovo: Joint Design of Protein Structure and Sequence Using a Unified Energy-based Model (ICML 2024 · similarity 0.87)
- Elucidating the Design Space of Multimodal Protein Language Models (ICML 2025 · similarity 0.87)
- FoldToken: Learning Protein Language via Vector Quantization and Beyond (AAAI 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.
