Machine Learning Prediction of Optical Properties of Coumarin Derivatives Using Gaussian-Weighted Graph Convolution and Subgraph Modular Input DOI
Seokwoo Kim, Minhi Han, Jin-Yong Park

et al.

Journal of Chemical Information and Modeling, Journal Year: 2025, Volume and Issue: unknown

Published: May 7, 2025

Coumarin derivatives have been widely developed and utilized as chromophores fluorophores in various research fields. In this study, we constructed an experimental database of the optical properties─specifically, absorption emission wavelengths measured solutions─and a machine learning (ML) model based on Gaussian-weighted graph convolution (GWGC) subgraph modular input (SMI) to predict these properties. The GWGC was introduced novel molecular representation that accounts for interatomic effects among neighboring atoms when properties coumarin were predicted. SMI represent subgraphs composed core six substituents, thereby modularizing vector into substituent vectors. This approach encodes both separate chemical information substituents well positional facilitating understanding how each influences core. ML models leveraging outperformed those RDKit descriptors count-based Morgan fingerprint. with can be generally applied molecules structure its substituents.

Language: Английский

Machine learning aided photovolatic property predictions, design and library generation of indeno-fluorene donors with lowest exciton bindings DOI
Hussein Ali Kadhim Kyhoiesh, Ashraf Y. Elnaggar,

Mustafa Al-Khafaji

et al.

Solar Energy, Journal Year: 2025, Volume and Issue: 291, P. 113399 - 113399

Published: March 6, 2025

Language: Английский

Citations

1
Machine Learning Prediction of Optical Properties of Coumarin Derivatives Using Gaussian-Weighted Graph Convolution and Subgraph Modular Input DOI
Seokwoo Kim, Minhi Han, Jin-Yong Park

et al.

Journal of Chemical Information and Modeling, Journal Year: 2025, Volume and Issue: unknown

Published: May 7, 2025

Coumarin derivatives have been widely developed and utilized as chromophores fluorophores in various research fields. In this study, we constructed an experimental database of the optical properties─specifically, absorption emission wavelengths measured solutions─and a machine learning (ML) model based on Gaussian-weighted graph convolution (GWGC) subgraph modular input (SMI) to predict these properties. The GWGC was introduced novel molecular representation that accounts for interatomic effects among neighboring atoms when properties coumarin were predicted. SMI represent subgraphs composed core six substituents, thereby modularizing vector into substituent vectors. This approach encodes both separate chemical information substituents well positional facilitating understanding how each influences core. ML models leveraging outperformed those RDKit descriptors count-based Morgan fingerprint. with can be generally applied molecules structure its substituents.

Language: Английский

Citations

0