《生命科学》 2025, 37(12): 1573-1586

基于人工智能大语言模型的小分子发现

杨雨薇¹ , 徐挺军^1,2 , 刘　蕾¹ , 吴顶峰¹ , 朱瑞新^1,*

¹同济大学生命科学与技术学院，上海市第一妇婴保健院转化医学中心，上海市母胎医学与妇科肿瘤研究 所，上海市母胎医学重点实验室，上海 200092 ²中国科学院上海有机化学研究所，上海 200032

摘　要：

小分子凭借其性能优势，在药物研发、材料创新、农业应用等领域展现出巨大潜力。人工智能，特别是大语言模型，在小分子生成、优化、表征、化学知识提取等小分子发现任务中有着出色的表现，探索了更加广阔的化学空间，显著提高了分子发现的高效性和精确性。本文概述了大语言模型在小分子发现领域中的应用，并讨论了当前面临的挑战以及发展趋势。

通讯作者：朱瑞新 , Email:rxzhu@tongji.edu.cn

Small molecule discovery based on artificial intelligence large language models

YANG Yu-Wei¹ , XU Ting-Jun^1,2 , LIU Lei¹ , WU Ding-Feng¹ , ZHU Rui-Xin^1,*

¹Shanghai Key Laboratory of Maternal Fetal Medicine, Shanghai Institute of Maternal-Fetal Medicine and Gynecologic Oncology, Clinical and Translation Research Center, Shanghai First Maternity and Infant Hospital, School of Life Sciences and Technology, Tongji University, Shanghai 200092, China ²Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032, China

Abstract:

Small molecules have demonstrated significant potential across various fields, such as pharmaceutical research, materials innovation, and agricultural applications, owing to their favorable properties. Artificial intelligence, particularly large language models, has shown remarkable capability in tasks related to small molecule generation, optimization, representation, and chemical knowledge extraction. This not only enables exploration of a broader chemical space, but also substantially improves the efficiency and accuracy of molecular discovery. This paper outlines the application of large language models in the field of small molecule discovery and discusses current challenges as well as future development trends.

Communication Author：ZHU Rui-Xin , Email:rxzhu@tongji.edu.cn