READRetro: natural product biosynthesis predicting with retrieval-augmented dual-view retrosynthesis

IF 8.3 1区生物学 Q1 PLANT SCIENCES

New Phytologist Pub Date : 2024-07-30 DOI:10.1111/nph.20012

Taein Kim, Seul Lee, Yejin Kwak, Min-Soo Choi, Jeongbin Park, Sung Ju Hwang, Sang-Gyu Kim

{"title":"READRetro: natural product biosynthesis predicting with retrieval-augmented dual-view retrosynthesis","authors":"Taein Kim, Seul Lee, Yejin Kwak, Min-Soo Choi, Jeongbin Park, Sung Ju Hwang, Sang-Gyu Kim","doi":"10.1111/nph.20012","DOIUrl":null,"url":null,"abstract":"<div>\n \n <p>\n \n </p><ul>\n \n \n <li>Plants, as a sessile organism, produce various secondary metabolites to interact with the environment. These chemicals have fascinated the plant science community because of their ecological significance and notable biological activity. However, predicting the complete biosynthetic pathways from target molecules to metabolic building blocks remains a challenge.</li>\n \n \n <li>Here, we propose retrieval-augmented dual-view retrosynthesis (READRetro) as a practical bio-retrosynthesis tool to predict the biosynthetic pathways of plant natural products. Conventional bio-retrosynthesis models have been limited in their ability to predict biosynthetic pathways for natural products. READRetro was optimized for the prediction of complex metabolic pathways by incorporating cutting-edge deep learning architectures, an ensemble approach, and two retrievers.</li>\n \n \n <li>Evaluation of single- and multi-step retrosynthesis showed that each component of READRetro significantly improved its ability to predict biosynthetic pathways. READRetro was also able to propose the known pathways of secondary metabolites such as monoterpene indole alkaloids and the unknown pathway of menisdaurilide, demonstrating its applicability to real-world bio-retrosynthesis of plant natural products.</li>\n \n \n <li>For researchers interested in the biosynthesis and production of secondary metabolites, a user-friendly website (https://readretro.net) and the open-source code of READRetro have been made available.</li>\n </ul>\n \n </div>","PeriodicalId":214,"journal":{"name":"New Phytologist","volume":null,"pages":null},"PeriodicalIF":8.3000,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"New Phytologist","FirstCategoryId":"99","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1111/nph.20012","RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"PLANT SCIENCES","Score":null,"Total":0}

引用次数: 0

Abstract

Plants, as a sessile organism, produce various secondary metabolites to interact with the environment. These chemicals have fascinated the plant science community because of their ecological significance and notable biological activity. However, predicting the complete biosynthetic pathways from target molecules to metabolic building blocks remains a challenge.
Here, we propose retrieval-augmented dual-view retrosynthesis (READRetro) as a practical bio-retrosynthesis tool to predict the biosynthetic pathways of plant natural products. Conventional bio-retrosynthesis models have been limited in their ability to predict biosynthetic pathways for natural products. READRetro was optimized for the prediction of complex metabolic pathways by incorporating cutting-edge deep learning architectures, an ensemble approach, and two retrievers.
Evaluation of single- and multi-step retrosynthesis showed that each component of READRetro significantly improved its ability to predict biosynthetic pathways. READRetro was also able to propose the known pathways of secondary metabolites such as monoterpene indole alkaloids and the unknown pathway of menisdaurilide, demonstrating its applicability to real-world bio-retrosynthesis of plant natural products.
For researchers interested in the biosynthesis and production of secondary metabolites, a user-friendly website (https://readretro.net) and the open-source code of READRetro have been made available.

查看原文本刊更多论文

READRetro：利用检索增强的双视角逆合成预测天然产物的生物合成。

植物作为一种无柄生物，会产生各种次生代谢物来与环境互动。这些化学物质具有重要的生态意义和显著的生物活性，因此吸引着植物科学界。然而，预测从目标分子到代谢构件的完整生物合成途径仍然是一项挑战。在此，我们提出了检索增强双视角逆合成（READRetro），作为预测植物天然产物生物合成途径的实用生物逆合成工具。传统的生物逆合成模型在预测天然产物的生物合成途径方面能力有限。READRetro 结合了前沿的深度学习架构、集合方法和两个检索器，针对复杂代谢途径的预测进行了优化。对单步和多步逆合成的评估表明，READRetro 的每个组件都显著提高了预测生物合成途径的能力。READRetro 还能提出单萜吲哚生物碱等次生代谢物的已知途径和 menisdaurilide 的未知途径，证明其适用于现实世界中植物天然产品的生物逆合成。对于对次生代谢物的生物合成和生产感兴趣的研究人员，我们提供了一个用户友好型网站（https://readretro.net）和 READRetro 的开放源代码。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

New Phytologist 生物-植物科学

自引率

5.30%

发文量

728

期刊介绍： New Phytologist is an international electronic journal published 24 times a year. It is owned by the New Phytologist Foundation, a non-profit-making charitable organization dedicated to promoting plant science. The journal publishes excellent, novel, rigorous, and timely research and scholarship in plant science and its applications. The articles cover topics in five sections: Physiology & Development, Environment, Interaction, Evolution, and Transformative Plant Biotechnology. These sections encompass intracellular processes, global environmental change, and encourage cross-disciplinary approaches. The journal recognizes the use of techniques from molecular and cell biology, functional genomics, modeling, and system-based approaches in plant science. Abstracting and Indexing Information for New Phytologist includes Academic Search, AgBiotech News & Information, Agroforestry Abstracts, Biochemistry & Biophysics Citation Index, Botanical Pesticides, CAB Abstracts®, Environment Index, Global Health, and Plant Breeding Abstracts, and others.