| Title: |
Enhancing atom mapping with multitask learning and symmetry-aware deep graph matching |
| Authors: |
Astero, Maryam; Rousu, Juho |
| Contributors: |
Department of Computer Science; Probabilistic Machine Learning; Computer Science Professors; Computer Science - Large-scale Computing and Data Analysis (LSCA) - Research area; Computer Science - Artificial Intelligence and Machine Learning (AIML) - Research area; Computer Science - Computational Life Sciences (CSLife) - Research area; Helsinki Institute for Information Technology (HIIT); Professorship Rousu Juho; Aalto-yliopisto; Aalto University |
| Publisher Information: |
BioMed Central |
| Publication Year: |
2025 |
| Collection: |
Aalto University Publication Archive (Aaltodoc) / Aalto-yliopiston julkaisuarkistoa |
| Subject Terms: |
Atom mapping; Graph matching; Graph representation learning; Multitask learning |
| Description: |
Publisher Copyright: © The Author(s) 2025. ; Atom mapping involves identifying the correspondence between individual atoms in reactant molecules and their counterparts in product molecules. This process is crucial for gaining deeper insight into reaction mechanisms, such as defining reaction templates and determining which chemical bonds are formed or broken during a reaction. However, reliable atom mapping data are often limited or incomplete within chemical databases, rendering manual annotation impractical for large-scale datasets. To address this limitation, we propose the Symmetry-Aware Multitask Atom Mapping Network (SAMMNet), a model designed to automatically infer atom correspondences by incorporating an auxiliary self-supervised task during training. SAMMNet employs molecular graph representations and leverages graph neural networks to capture both general and task-specific features, enabling enhanced predictive performance. Our experimental results demonstrate that the multitask learning framework, coupled with symmetry-aware atom mapping, improves accuracy and robustness in atom mapping predictions. This makes our method a promising advancement for computational chemistry and related fields. ; Peer reviewed |
| Document Type: |
article in journal/newspaper |
| File Description: |
application/pdf |
| Language: |
English |
| Relation: |
Journal of Cheminformatics; Volume 17, issue 1, pp. 1-17; M.A. acknowledges Elena Casiraghi for her invaluable assistance in reviewing this work and providing constructive feedback. We acknowledge the generous support from the Wihuri Foundation as well as the Jane and Aatos Erkko Foundation (BIODESIGN project), which contributed to the advancement of this study. Additionally, this research has in part been funded by the Research Council of Finland (Grants 339421 and 345802).; https://aaltodoc.aalto.fi/handle/123456789/136782 |
| DOI: |
10.1186/s13321-025-01030-3 |
| Availability: |
https://aaltodoc.aalto.fi/handle/123456789/136782; https://doi.org/10.1186/s13321-025-01030-3 |
| Rights: |
openAccess ; CC BY-NC-ND ; https://creativecommons.org/licenses/by-nc-nd/4.0/ |
| Accession Number: |
edsbas.3BFC8823 |
| Database: |
BASE |