Research map / Large language models / Agents, code and math
PDAgent: An LLM-Driven Autonomous Agent Framework Towards *In Silico* Protein Design via Directed Mutation
Song Ouyang, Zhijie Dong, Yong Luo, Kehua Su, Huangxuan Zhao, Miaojing Shi, Bo Du
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 the iteration loop (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | average CSR: 81.85 -> 90.60 | Improved |
| add history memory (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | average CSR: 89.36 -> 90.60 | Improved |
| separate THINK and PLAN phases (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | average CSR: 86.53 -> 90.60 | Improved |
| add conservation guard (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | average CSR: 89.65 -> 90.60 | Improved |
| add conservation guard (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | foldability: 88.33 -> 87.69 | Hurt |
| add mutation loop to template-only output (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | average CSR: 82.24 -> 90.60 | Improved |
| use knowledge-guided mutation instead of random generation (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | average CSR: 55.46 -> 90.60 | Improved |
| use knowledge-guided mutation instead of random generation (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | foldability: 30.24 -> 87.69 | Improved |
| increase candidate templates from 1 to 3 (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks | CSR: 72.5 -> 90.6 | Improved |
| increase candidate templates from 1 to 3 (ablation) | PDAgent (Qwen-3-30B) · 100 design tasks | CSR: 67.3 -> 78.9 | Improved |
| set conservation threshold to 0.6 rather than 0.2 or 0.8 (ablation) | PDAgent (LLaMA-3.1-8B) · 100 design tasks, n=3 | CSR and pLDDT | Improved |
| replace best direct-LLM baseline Gemini-3-Pro with PDAgent | PDAgent (DeepSeek-V3.2) · 100 design tasks | average CSR: 82.55 -> 91.86 | Improved |
Closest papers by meaning
- Protein Design with Agent Rosetta: A Case Study for Specialized Scientific Agents (ICML 2026 · similarity 0.92)
- A Variational Perspective on Generative Protein Fitness Optimization (ICML 2025 · similarity 0.88)
- ProteinOPD: Towards Effective and Efficient Preference Alignment for Protein Design (NeurIPS 2026 · similarity 0.87)
- Probe Before You Edit: Probing-Guided Molecular Optimization for LLM Agents in Structure-Based Drug Design (NeurIPS 2026 · similarity 0.86)
- Knowledge-aware Reinforced Language Models for Protein Directed Evolution (ICML 2024 · similarity 0.86)
- pCoMole: Pareto-Constrained Molecule Editing with Discrete Flows (NeurIPS 2026 · similarity 0.85)
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.
