TY - JOUR

T1 - Fine-Tuning Directional Message Passing Neural Networks: Predicting Properties of Conjugated Organic Polymers with High Accuracy

AU - Koskin, Igor P.

AU - Petrosyan, Lev S.

AU - Kazantsev, Maxim S.

N1 - This work was supported by RSF project 25-73-00401 (https://rscf.ru/project/25-73-00401/ accessed on 30 March 2026).

PY - 2026/4/2

Y1 - 2026/4/2

N2 - Conjugated organic polymers are the cornerstone of modern organic electronics, yet accurate prediction of their properties remains a challenging task due to their synthetic complexity and high computational cost of quantum-chemical methods. Here, we develop a graph neural network based on the DimeNet++ direct message passing architecture to predict HOMO, LUMO and energy gaps of conjugated polymers directly from their 3D monomer structure. The model was pre-trained on TD-DFT-extrapolated data and trained on a limited dataset of experimentally measured properties. As a result, pre-training had significantly improved model’s accuracy compared to direct training (MAEs ~0.3 eV vs. 0.074 eV, 0.141 and 0.172 for HOMO/LUMO and energy gap, respectively). Pre-training on monomer DFT data did not provide comparable gains. The results demonstrate that polymer-relevant pre-training is critical for capturing structure–property relationships and enables accurate predictions without delta-learning or prior quantum-chemical calculations, facilitating efficient screening and rational design of conjugated polymers for organic optoelectronics.

AB - Conjugated organic polymers are the cornerstone of modern organic electronics, yet accurate prediction of their properties remains a challenging task due to their synthetic complexity and high computational cost of quantum-chemical methods. Here, we develop a graph neural network based on the DimeNet++ direct message passing architecture to predict HOMO, LUMO and energy gaps of conjugated polymers directly from their 3D monomer structure. The model was pre-trained on TD-DFT-extrapolated data and trained on a limited dataset of experimentally measured properties. As a result, pre-training had significantly improved model’s accuracy compared to direct training (MAEs ~0.3 eV vs. 0.074 eV, 0.141 and 0.172 for HOMO/LUMO and energy gap, respectively). Pre-training on monomer DFT data did not provide comparable gains. The results demonstrate that polymer-relevant pre-training is critical for capturing structure–property relationships and enables accurate predictions without delta-learning or prior quantum-chemical calculations, facilitating efficient screening and rational design of conjugated polymers for organic optoelectronics.

KW - directional message passing

KW - electronic property prediction

KW - fine-tuning

KW - graph neural network

KW - learning transfer

KW - polymer informatics

UR - https://www.scopus.com/pages/publications/105035517202

UR - https://www.mendeley.com/catalogue/61bc6751-0bca-3ae1-b73c-69b408238771/

U2 - 10.3390/polym18070879

DO - 10.3390/polym18070879

M3 - Article

VL - 18

JO - Polymers

JF - Polymers

SN - 2073-4360

IS - 7

M1 - 879

ER -