Retroplanner improves chemical synthesis plans by comparing multiple routes
RetroPlanner: Multi-Route Expansion and Multi-Objective Reasoning for LLM-Based Retrosynthesis Planning
Artificial IntelligenceComputation and LanguageMachine Learning
Summary
Planning how to make molecules in the lab can be tricky because there are many ways to build them, but not all are practical or affordable. The authors created RetroPlanner, which uses artificial intelligence to look at several possible ways to make a molecule at the same time and decide which routes are better by considering how plausible the reactions are, how consistent they are, and how much the starting materials cost. This approach helps find more alternative synthesis routes and better choices than older methods that only considered one route at a time.
What this means in practice
- •For chemical engineers: Generate multiple practical synthesis routes for target molecules, optimizing for cost and reaction reliability.
- •For pharmaceutical developers: Design candidate synthetic pathways to drug molecules that balance feasibility and cost using AI-assisted route comparison.
Authors
Ye Rin Kim, Yongjun Jeong, Seongjun Yun, Hanbum Ko, Soohyun Oh, Sungbin Lim, Sungwoong Kim
Abstract
Multi-step retrosynthesis planning is often considered solved once a route to purchasable starting materials is found, but such a route should also be practical in real-world chemistry. Recent LLM-based planners formulate this process as sequential decision-making over an AND-OR graph, yet typically expand a single molecule per turn and are optimized for solvability alone, leaving alternative routes uncompared across the multiple objectives that determine practicality before the search budget is committed. We introduce RetroPlanner, an LLM-based planner that jointly compares multiple routes to steer the search toward higher synthesizability, as measured by reaction plausibility, round-trip consistency, and starting-material price. RetroPlanner is trained to reason over chemical evidence: structural progress from the target molecule, bond changes and reaction classes, where candidate routes diverge from one another, and how they rank across multiple objectives. Experiments on multi-step retrosynthesis benchmarks demonstrate that RetroPlanner recovers more alternative routes than prior LLM-based planners while improving synthetic route quality and maintaining a high success rate.