Research map / Probabilistic and scientific modeling / Molecules and proteins
PDFBench: A Benchmark for De Novo Protein Design from Function
Jiahao Kuang, Nuowei Liu, Changzhi Sun, Jie Wang, Tao Ji, Yuanbin Wu
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 |
|---|---|---|---|
| compare ProDVa against Pinal on foldability | ProDVa · Description-guided (MolinstTest) | pLDDT: 75.25 -> 76.86 | Improved |
| compare ProDVa against Pinal on retrieval accuracy | ProDVa · Description-guided (MolinstTest) | Retrieval Accuracy: 63.43 -> 66.83 | Improved |
| use ProDVa instead of ProteinDT | ProDVa · Description-guided (MolinstTest) | Retrieval Accuracy: 16.92 -> 66.83 | Improved |
| use ProDVa instead of Pinal on novelty (sequence) | ProDVa · Description-guided (MolinstTest) | Novelty-Seq: 43.82 -> 14.64 | Hurt |
| use ProDVa instead of Pinal on diversity (sequence) | ProDVa · Description-guided (MolinstTest) | Diversity-Seq: 75.88 -> 46.10 | Hurt |
| use CFP-Gen instead of ProDVa on perplexity | CFP-Gen · Keyword-guided (SwissTest, IPR keywords) | Repeat: 1.99 -> 11.86 | Hurt |
| change keyword guidance from GO to IPR&GO | ProDVa · Keyword-guided (SwissTest) | Retrieval Accuracy: 52.38 -> 61.23 | Improved |
| change keyword guidance from GO to IPR&GO | CFP-Gen · Keyword-guided (SwissTest) | Retrieval Accuracy: 38.87 -> 45.77 | Improved |
| hard instead of soft retrieval negatives for retrieval accuracy | Retrieval Accuracy metric · Description-guided (MolinstTest) | Retrieval Accuracy: 60.64% gap (ProDVa) | Hurt |
| Repeat above 10 in designed sequences | Designed proteins · Keyword-guided (SwissTest) | pLDDT | Hurt |
Closest papers by meaning
- ProteinBench: A Holistic Evaluation of Protein Foundation Models (ICLR 2025 · similarity 0.90)
- CarbonNovo: Joint Design of Protein Structure and Sequence Using a Unified Energy-based Model (ICML 2024 · similarity 0.87)
- Generative Enzyme Design Guided by Functionally Important Sites and Small-Molecule Substrates (ICML 2024 · similarity 0.86)
- CBGBench: Fill in the Blank of Protein-Molecule Complex Binding Graph (ICLR 2025 · similarity 0.86)
- Protein Design with Dynamic Protein Vocabulary (NeurIPS 2025 · similarity 0.86)
- PSBench: a large-scale benchmark for estimating the accuracy of protein complex structural models (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.
