Research

Paper

AI LLM March 13, 2026

RetroReasoner: A Reasoning LLM for Strategic Retrosynthesis Prediction

Authors

Hanbum Ko, Chanhui Lee, Ye Rin Kim, Rodrigo Hormazabal, Sehui Han, Sungbin Lim, Sungwoong Kim

Abstract

Retrosynthesis prediction is a core task in organic synthesis that aims to predict reactants for a given product molecule. Traditionally, chemists select a plausible bond disconnection and derive corresponding reactants, which is time-consuming and requires substantial expertise. While recent advancements in molecular large language models (LLMs) have made progress, many methods either predict reactants without strategic reasoning or conduct only a generic product analysis, rather than reason explicitly about bond-disconnection strategies that logically lead to the choice of specific reactants. To overcome these limitations, we propose RetroReasoner, a retrosynthetic reasoning model that leverages chemists' strategic thinking. RetroReasoner is trained using both supervised fine-tuning (SFT) and reinforcement learning (RL). For SFT, we introduce SyntheticRetro, a framework that generates structured disconnection rationales alongside reactant predictions. In the case of RL, we apply a round-trip accuracy as reward, where predicted reactants are passed through a forward synthesis model, and predictions are rewarded when the forward-predicted product matches the original input product. Experimental results show that RetroReasoner not only outperforms prior baselines but also generates a broader range of feasible reactant proposals, particularly in handling more challenging reaction instances.

Metadata

arXiv ID: 2603.12666
Provider: ARXIV
Primary Category: cs.LG
Published: 2026-03-13
Fetched: 2026-03-16 06:01

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.12666v1</id>\n    <title>RetroReasoner: A Reasoning LLM for Strategic Retrosynthesis Prediction</title>\n    <updated>2026-03-13T05:20:56Z</updated>\n    <link href='https://arxiv.org/abs/2603.12666v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.12666v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Retrosynthesis prediction is a core task in organic synthesis that aims to predict reactants for a given product molecule. Traditionally, chemists select a plausible bond disconnection and derive corresponding reactants, which is time-consuming and requires substantial expertise. While recent advancements in molecular large language models (LLMs) have made progress, many methods either predict reactants without strategic reasoning or conduct only a generic product analysis, rather than reason explicitly about bond-disconnection strategies that logically lead to the choice of specific reactants. To overcome these limitations, we propose RetroReasoner, a retrosynthetic reasoning model that leverages chemists' strategic thinking. RetroReasoner is trained using both supervised fine-tuning (SFT) and reinforcement learning (RL). For SFT, we introduce SyntheticRetro, a framework that generates structured disconnection rationales alongside reactant predictions. In the case of RL, we apply a round-trip accuracy as reward, where predicted reactants are passed through a forward synthesis model, and predictions are rewarded when the forward-predicted product matches the original input product. Experimental results show that RetroReasoner not only outperforms prior baselines but also generates a broader range of feasible reactant proposals, particularly in handling more challenging reaction instances.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-03-13T05:20:56Z</published>\n    <arxiv:comment>26 pages, 18 figures</arxiv:comment>\n    <arxiv:primary_category term='cs.LG'/>\n    <author>\n      <name>Hanbum Ko</name>\n    </author>\n    <author>\n      <name>Chanhui Lee</name>\n    </author>\n    <author>\n      <name>Ye Rin Kim</name>\n    </author>\n    <author>\n      <name>Rodrigo Hormazabal</name>\n    </author>\n    <author>\n      <name>Sehui Han</name>\n    </author>\n    <author>\n      <name>Sungbin Lim</name>\n    </author>\n    <author>\n      <name>Sungwoong Kim</name>\n    </author>\n  </entry>"
}