首都师范大学学报（自然科学版）2026，Vol.47Issue(3)：61-73,13.DOI:10.19789/j.1004-9398.2026.03.005
生成式AI与逆向设计:高通量预测新型荧光染料的分子结构
Generative AI for inverse design:high-throughput discovery of fluorescent dye molecules
摘要
Abstract
Fluorescent materials have been recognized as a series of important luminescent materials,which has important application in display,bioimaging,and chemical sensing.Conventional synthesis of fluorescent materials often relied on the trial-and-error method,suppressing the development of this realm and prolonging the development cycle.Generally,the recent advances in artificial intelligence(AI)technology have brought new opportunities to chemistry,furnishing the transformation of fluorescent molecular design from experience-based to data-driven approach.In this context,several key questions arise regarding A᷄I-empowered design and development of novel fluorescent dyes.What generative models and training strategies are applicable?What high-quality datasets are commonly used?How should molecular descriptors be selected?What are representative examples of inverse-design strategies?To address these questions,this article first outlines the fundamental principles of generative AI and inverse design.Starting from the key step of complex material development-molecular design,we summarize several commonly used generative models for molecular design and compare their strengths and limitations.Subsequently,we introduce the prediction of key properties of fluorescent dyes based on commonly employed fluorescent dye databases and molecular descriptors.Finally,we review domestic and international progress in applying generative AI to fluorescent dye design and discuss potential future directions.By focusing on actionable research pathways,we aim to provide a reference for establishing AI as a reliable engine driving the design and discovery of fluorescent dyes and functional molecules.关键词
生成式AI/荧光染料/逆向设计/机器学习/分子设计Key words
generative AI/fluorescent dye/inverse design/machine learning/molecular design分类
化学化工引用本文复制引用
彭灵雅,窦小雅,刘乐洁,江艳..生成式AI与逆向设计:高通量预测新型荧光染料的分子结构[J].首都师范大学学报（自然科学版）,2026,47(3):61-73,13.基金项目
国家自然科学基金项目(22403061) （22403061）
陕西省科协青年人才计划托举计划项目(20250608) （20250608）
中央高校基础科研业务费(GK202406028) （GK202406028）
