Boosting the Prediction Accuracy of Glass Transition Temperature in Polyimides: A Hybrid Machine Learning Approach Integrating Morgan Fingerprints and Molecular Descriptors
Molecular Informatics · 2026-08 · 45 (8)
DOI: 10.1002/minf.70049 ↗Kurz gefasst
Fachartikel in der Zeitschrift Molecular Informatics, online erschienen am 23.08.2026. Verfasst von Peishuai Xing und 8 weiteren Autor:innen. Beteiligte Einrichtung: School of Electronic Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China. Hauptthemen laut MEDLINE: Machine Learning, Transition Temperature, Imides, Glass. Die Arbeit nennt 44 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
The glass transition temperature ( T g ) of polyimides is a critical parameter determining their processability and application performance. Traditional experimental methods for measuring T g are time‐consuming and costly, while existing machine learning prediction models predominantly rely on manually defined molecular descriptors, which often fail to fully capture detailed molecular structural information, limiting their prediction accuracy and generalization capability. To address this, this study proposes a hybrid feature engineering strategy combining Morgan fingerprints and molecular descriptors to comprehensively represent the chemical structure of polyimides. Based on a dataset of 1257 polyimide samples from a public database, we systematically compared six feature selection methods and employed multiple mainstream machine learning algorithms for modeling. The results show that the CATB model performed best, achieving a coefficient of determination ( R 2 ) of 0.882 and a mean absolute error (MAE) of 17.34 °C on an independent test set, with fivefold cross‐validation further confirming the model's robustness. SHAP interpretability analysis revealed the significant influence of key features such as the number of rotatable bonds, ether bonds, and ether‐linked oxyethylene units on T g , providing clear guidance for molecular design. External validation demonstrated the model's strong generalization ability. This study not only achieves high‐precision and robust T g prediction but also highlights the importance of hybrid feature strategies in polymer property modeling, offering a data‐driven foundation for the rational design of polyimides.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- School of Electronic Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China.
- School of Mechanical Engineering, XinJiang University, Urumchi, China.
- Semiconductor Technology Research Department, Ji Hua Laboratory, Foshan, China.
- South China Normal University, Guangzhou, China.
ORCID-Profile: Peishuai Xing
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Xing, P., Guo, X., Wang, Y., Chen, Z., Shen, S., He, B., Li, T., Peng, D., & Yang, Z. (2026). Boosting the Prediction Accuracy of Glass Transition Temperature in Polyimides: A Hybrid Machine Learning Approach Integrating Morgan Fingerprints and Molecular Descriptors. Molecular Informatics, 45(8). https://doi.org/10.1002/minf.70049
Vancouver
Xing P, Guo X, Wang Y, Chen Z, Shen S, He B, et al. Boosting the Prediction Accuracy of Glass Transition Temperature in Polyimides: A Hybrid Machine Learning Approach Integrating Morgan Fingerprints and Molecular Descriptors. Molecular Informatics. 2026;45(8). doi:10.1002/minf.70049
BibTeX
@article{Xing2026,
author = {Xing, Peishuai and Guo, Xiaodong and Wang, Yang and Chen, Zeting and Shen, Shicheng and He, Bin and Li, Tianyun and Peng, Dongmei and Yang, Zhenhuai},
title = {Boosting the Prediction Accuracy of Glass Transition Temperature in Polyimides: A Hybrid Machine Learning Approach Integrating Morgan Fingerprints and Molecular Descriptors},
journal = {Molecular Informatics},
year = {2026},
volume = {45},
number = {8},
doi = {10.1002/minf.70049}
}
EUVIMED