MMORF: A Multi-agent Framework for Designing Multi-objective Retrosynthesis Planning Systems

2026-04-06

AI Summary

Multi-objective retrosynthetic planning requires dynamic trade-offs among conflicting objectives such as product quality, safety, and cost. This work proposes MMORF, the first modular multi-agent framework for this task, which leverages composable, specialized agents to enable flexible configuration and systematic evaluation under both soft and hard constraints. By integrating large language models, modular components, and Pareto optimization, MMORF achieves efficient pathway search. Evaluated on a new benchmark comprising 218 tasks, the proposed approach significantly outperforms existing baselines: MASIL achieves state-of-the-art performance on soft-constraint tasks, while RFAS surpasses the current best method on hard-constraint tasks with a success rate of 48.6%.

Abstract

Multi-objective retrosynthesis planning is a critical chemistry task requiring dynamic balancing of quality, safety, and cost objectives. Language model-based multi-agent systems (MAS) offer a promising approach for this task: leveraging interactions of specialized agents to incorporate multiple objectives into retrosynthesis planning. We present MMORF, a framework for constructing MAS for multi-objective retrosynthesis planning. MMORF features modular agentic components, which can be flexibly combined and configured into different systems, enabling principled evaluation and comparison of different system designs. Using MMORF, we construct two representative MAS: MASIL and RFAS. On a newly curated benchmark consisting of 218 multi-objective retrosynthesis planning tasks, MASIL achieves strong safety and cost metrics on soft-constraint tasks, frequently Pareto-dominating baseline routes, while RFAS achieves a 48.6% success rate on hard-constraint tasks, outperforming state-of-the-art baselines. Together, these results show the effectiveness of MMORF as a foundational framework for exploring MAS for multi-objective retrosynthesis planning. Code and data are available at https://anonymous.4open.science/r/MMORF/.

Problem

• multi-objective retrosynthesis planning
• chemical synthesis
• safety
• cost
• quality

Innovation

• multi-agent system
• modular framework
• multi-objective retrosynthesis
• Pareto dominance
• constraint-aware planning