eprintid: 10196206 rev_number: 9 eprint_status: archive userid: 699 dir: disk0/10/19/62/06 datestamp: 2024-08-28 09:42:57 lastmod: 2024-08-28 09:42:57 status_changed: 2024-08-28 09:42:57 type: article metadata_visibility: show sword_depositor: 699 creators_name: Anselmi, Marco creators_name: Slabaugh, Greg creators_name: Crespo-Otero, Rachel creators_name: Di Tommaso, Devis title: Molecular graph transformer: stepping beyond ALIGNN into long-range interactions ispublished: pub divisions: UCL divisions: B04 divisions: C06 divisions: F56 keywords: Science & Technology, Physical Sciences, Technology, Chemistry, Multidisciplinary, Computer Science, Interdisciplinary Applications, Chemistry, Computer Science, PREDICTION, FRAMEWORK note: This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. abstract: Graph Neural Networks (GNNs) have revolutionized material property prediction by learning directly from the structural information of molecules and materials. However, conventional GNN models rely solely on local atomic interactions, such as bond lengths and angles, neglecting crucial long-range electrostatic forces that affect certain properties. To address this, we introduce the Molecular Graph Transformer (MGT), a novel GNN architecture that combines local attention mechanisms with message passing on both bond graphs and their line graphs, explicitly capturing long-range interactions. Benchmarking on MatBench and Quantum MOF (QMOF) datasets demonstrates that MGT's improved understanding of electrostatic interactions significantly enhances the prediction accuracy of properties like exfoliation energy and refractive index, while maintaining state-of-the-art performance on all other properties. This breakthrough paves the way for the development of highly accurate and efficient materials design tools across diverse applications. date: 2024-05-01 date_type: published publisher: ROYAL SOC CHEMISTRY official_url: http://dx.doi.org/10.1039/d4dd00014e oa_status: green full_text_type: pub language: eng primo: open primo_central: open_green verified: verified_manual elements_id: 2278039 doi: 10.1039/d4dd00014e lyricists_name: Crespo Otero, Rachel lyricists_id: RCRES48 actors_name: Crespo Otero, Rachel actors_id: RCRES48 actors_role: owner funding_acknowledgements: [Mini-centre-for-Doctoral-Training in CO2-conversion at QMUL for PhD scholarship]; [Hartree Center at the University of Oxford]; EP/T022108/1 [EPSRC]; EP/T022205/1 [EPSRC]; [HPC Midlands Plus] full_text_status: public publication: Digital Discovery volume: 3 number: 5 pagerange: 1048-1057 pages: 10 issn: 2635-098X citation: Anselmi, Marco; Slabaugh, Greg; Crespo-Otero, Rachel; Di Tommaso, Devis; (2024) Molecular graph transformer: stepping beyond ALIGNN into long-range interactions. Digital Discovery , 3 (5) pp. 1048-1057. 10.1039/d4dd00014e <https://doi.org/10.1039/d4dd00014e>. Green open access document_url: https://discovery.ucl.ac.uk/id/eprint/10196206/1/Crespo%20Otero_d4dd00014e.pdf