@inproceedings{liu-etal-2026-kgrxn,
title = "{KGR}xn-{LLM}: Knowledge Graph Enhanced Large Language Models for Molecular Reaction Reasoning",
author = "Liu, Weichen and
Xue, Qiyao and
Wu, Yuyang and
Isayev, Olexandr and
Miskov-Zivanov, Natasa",
editor = "Demner-Fushman, Dina and
Ananiadou, Sophia and
Roberts, Kirk and
Tsujii, Junichi",
booktitle = "{B}io{NLP} 2026",
month = jul,
year = "2026",
address = "San Diego, California",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.bionlp-1.22/",
pages = "268--281",
ISBN = "979-8-89176-434-7",
abstract = "Large language models (LLMs) demonstrate strong general language capabilities but remain limited in chemical reasoning, particularly for tasks requiring structured, mechanistic understanding of molecular reactions. We present Knowledge Graph Reaction LLM (KGRxn-LLM), a framework that augments LLMs with a hierarchical chemical knowledge graph (KG) to ground reasoning in molecular transformations and reaction patterns. Existing benchmarks primarily emphasize reaction or molecular fact recall, providing limited assessment of reaction-level mechanistic reasoning. To address this gap, we introduce KGRxn-Bench, a benchmark of 1,200 questions designed to evaluate LLMs on reaction-centric reasoning tasks, including functional group identification, reaction type classification, and product and reagent prediction. Experimental results show that our approach of grounding LLMs in structured KG substantially improves performance across multiple tasks and model backbones, outperforming domain-specific fine-tuned models on KG-covered splits and most hold-out splits."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="liu-etal-2026-kgrxn">
<titleInfo>
<title>KGRxn-LLM: Knowledge Graph Enhanced Large Language Models for Molecular Reaction Reasoning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Weichen</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qiyao</namePart>
<namePart type="family">Xue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuyang</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Olexandr</namePart>
<namePart type="family">Isayev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Natasa</namePart>
<namePart type="family">Miskov-Zivanov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>BioNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dina</namePart>
<namePart type="family">Demner-Fushman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sophia</namePart>
<namePart type="family">Ananiadou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kirk</namePart>
<namePart type="family">Roberts</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junichi</namePart>
<namePart type="family">Tsujii</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-434-7</identifier>
</relatedItem>
<abstract>Large language models (LLMs) demonstrate strong general language capabilities but remain limited in chemical reasoning, particularly for tasks requiring structured, mechanistic understanding of molecular reactions. We present Knowledge Graph Reaction LLM (KGRxn-LLM), a framework that augments LLMs with a hierarchical chemical knowledge graph (KG) to ground reasoning in molecular transformations and reaction patterns. Existing benchmarks primarily emphasize reaction or molecular fact recall, providing limited assessment of reaction-level mechanistic reasoning. To address this gap, we introduce KGRxn-Bench, a benchmark of 1,200 questions designed to evaluate LLMs on reaction-centric reasoning tasks, including functional group identification, reaction type classification, and product and reagent prediction. Experimental results show that our approach of grounding LLMs in structured KG substantially improves performance across multiple tasks and model backbones, outperforming domain-specific fine-tuned models on KG-covered splits and most hold-out splits.</abstract>
<identifier type="citekey">liu-etal-2026-kgrxn</identifier>
<location>
<url>https://aclanthology.org/2026.bionlp-1.22/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>268</start>
<end>281</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T KGRxn-LLM: Knowledge Graph Enhanced Large Language Models for Molecular Reaction Reasoning
%A Liu, Weichen
%A Xue, Qiyao
%A Wu, Yuyang
%A Isayev, Olexandr
%A Miskov-Zivanov, Natasa
%Y Demner-Fushman, Dina
%Y Ananiadou, Sophia
%Y Roberts, Kirk
%Y Tsujii, Junichi
%S BioNLP 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California
%@ 979-8-89176-434-7
%F liu-etal-2026-kgrxn
%X Large language models (LLMs) demonstrate strong general language capabilities but remain limited in chemical reasoning, particularly for tasks requiring structured, mechanistic understanding of molecular reactions. We present Knowledge Graph Reaction LLM (KGRxn-LLM), a framework that augments LLMs with a hierarchical chemical knowledge graph (KG) to ground reasoning in molecular transformations and reaction patterns. Existing benchmarks primarily emphasize reaction or molecular fact recall, providing limited assessment of reaction-level mechanistic reasoning. To address this gap, we introduce KGRxn-Bench, a benchmark of 1,200 questions designed to evaluate LLMs on reaction-centric reasoning tasks, including functional group identification, reaction type classification, and product and reagent prediction. Experimental results show that our approach of grounding LLMs in structured KG substantially improves performance across multiple tasks and model backbones, outperforming domain-specific fine-tuned models on KG-covered splits and most hold-out splits.
%U https://aclanthology.org/2026.bionlp-1.22/
%P 268-281
Markdown (Informal)
[KGRxn-LLM: Knowledge Graph Enhanced Large Language Models for Molecular Reaction Reasoning](https://aclanthology.org/2026.bionlp-1.22/) (Liu et al., BioNLP 2026)
ACL