Research map / Multi-agent systems and games / Cooperative multi-agent RL
Advancing Protein Design via Multi-Agent Reinforcement Learning with Pareto-Based Collaborative Optimization
Mingming Zhu, Jiahua Rao, Xiaoyu Chen, Qianmu Yuan, Yuedong Yang
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 consensus-weighted negotiation on top of multi-agent DPO (ablation) | MAProt · Megascale | success rate (%): 82.9 -> 89.7 | Improved |
| add consensus-weighted negotiation on top of multi-agent DPO (ablation) | MAProt · Megascale | median scRMSD: 1.012 -> 0.890 | Improved |
| add consensus negotiation to multi-agent DPO (ablation) | MAProt · Megascale | median Pred-ddG: 2.143 -> 1.603 | Hurt |
| use multiple agents instead of ProteinMPNN alone (DPO) (ablation) | ProteinMPNN · Megascale | success rate (%): 79.1 -> 82.9 | Improved |
| use multiple agents instead of ProteinMPNN alone (SFT) (ablation) | ProteinMPNN · Megascale | success rate (%): 66.7 -> 69.8 | Improved |
| add SaProt agent to ProteinMPNN under DPO (ablation) | ProteinMPNN · Megascale | success rate (%): 79.1 -> 84.1 | Improved |
| add all three agents rather than ProteinMPNN+SaProt under DPO (ablation) | MAProt (DPO) · Megascale | success rate (%): 84.1 -> 82.9 | Hurt |
| multi-agent zero-shot vs ProteinMPNN zero-shot (ablation) | MAProt · Megascale | success rate (%): 31.5 -> 54.4 | Improved |
| DPO fine-tuning vs SFT for ProteinMPNN (ablation) | ProteinMPNN · Megascale | success rate (%): 66.7 -> 79.1 | Improved |
| lower beta to 0.1 vs 0.5 (ablation) | MAProt · Megascale | structural deviation | Hurt |
| raise beta to 1.0 vs 0.5 (ablation) | MAProt · Megascale | fitness / success rate | Hurt |
| add consensus negotiation to multi-agent DPO (ablation) | MAProt · AffinityDesign | success (%): 22.3 -> 32.1 | Improved |
Closest papers by meaning
- ProteinOPD: Towards Effective and Efficient Preference Alignment for Protein Design (NeurIPS 2026 · similarity 0.90)
- Protein Design with Agent Rosetta: A Case Study for Specialized Scientific Agents (ICML 2026 · similarity 0.89)
- Knowledge-aware Reinforced Language Models for Protein Directed Evolution (ICML 2024 · similarity 0.88)
- Multi-Objective Molecular Design Through Learning Latent Pareto Set (AAAI 2025 · similarity 0.87)
- ProSpero: Active Learning for Robust Protein Design Beyond Wild-Type Neighborhoods (NeurIPS 2025 · similarity 0.87)
- DePLM: Denoising Protein Language Models for Property Optimization (NeurIPS 2024 · 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.
