Interpretable prediction of aggregation-induced emission molecules based on graph neural networks

Shi Chen Zhang; Jun Zhu; Yi Zeng; Hua Qi Mai; Dong Wang; Xiao Yan Zheng

doi:10.1039/d5cc01949d

Interpretable prediction of aggregation-induced emission molecules based on graph neural networks

Shi Chen Zhang, Jun Zhu, Yi Zeng, Hua Qi Mai, Dong Wang^*, Xiao Yan Zheng^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

Abstract

We developed an interpretable graph neural network (96.4% accuracy) for AIEgen identification, revealing 24 characteristic functional groups. Based on these insights, two virtual library strategies (self-fragment and donor-acceptor docking) were proposed and predicted four experimentally confirmed AIEgens successfully, which establishes a rational design framework for AIE materials.

Original language	English
Journal	Chemical Communications
DOIs	https://doi.org/10.1039/d5cc01949d
Publication status	Accepted/In press - 2025
Externally published	Yes

Access to Document

10.1039/d5cc01949d

Cite this

@article{0cb8b596e70d410490df77a9c6f56766,

title = "Interpretable prediction of aggregation-induced emission molecules based on graph neural networks",

abstract = "We developed an interpretable graph neural network (96.4% accuracy) for AIEgen identification, revealing 24 characteristic functional groups. Based on these insights, two virtual library strategies (self-fragment and donor-acceptor docking) were proposed and predicted four experimentally confirmed AIEgens successfully, which establishes a rational design framework for AIE materials.",

author = "Zhang, {Shi Chen} and Jun Zhu and Yi Zeng and Mai, {Hua Qi} and Dong Wang and Zheng, {Xiao Yan}",

note = "Publisher Copyright: {\textcopyright} 2025 The Royal Society of Chemistry.",

year = "2025",

doi = "10.1039/d5cc01949d",

language = "English",

journal = "Chemical Communications",

issn = "1359-7345",

publisher = "Royal Society of Chemistry",

}

TY - JOUR

T1 - Interpretable prediction of aggregation-induced emission molecules based on graph neural networks

AU - Zhang, Shi Chen

AU - Zhu, Jun

AU - Zeng, Yi

AU - Mai, Hua Qi

AU - Wang, Dong

AU - Zheng, Xiao Yan

PY - 2025

Y1 - 2025

N2 - We developed an interpretable graph neural network (96.4% accuracy) for AIEgen identification, revealing 24 characteristic functional groups. Based on these insights, two virtual library strategies (self-fragment and donor-acceptor docking) were proposed and predicted four experimentally confirmed AIEgens successfully, which establishes a rational design framework for AIE materials.

AB - We developed an interpretable graph neural network (96.4% accuracy) for AIEgen identification, revealing 24 characteristic functional groups. Based on these insights, two virtual library strategies (self-fragment and donor-acceptor docking) were proposed and predicted four experimentally confirmed AIEgens successfully, which establishes a rational design framework for AIE materials.

UR - http://www.scopus.com/inward/record.url?scp=105005891963&partnerID=8YFLogxK

U2 - 10.1039/d5cc01949d

DO - 10.1039/d5cc01949d

M3 - Article

AN - SCOPUS:105005891963

SN - 1359-7345

JO - Chemical Communications

JF - Chemical Communications

ER -

Interpretable prediction of aggregation-induced emission molecules based on graph neural networks

Abstract

Access to Document

Other files and links

Fingerprint

Cite this