{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T20:31:58Z","timestamp":1772137918681,"version":"3.50.1"},"reference-count":43,"publisher":"IOP Publishing","issue":"2","license":[{"start":{"date-parts":[[2025,4,14]],"date-time":"2025-04-14T00:00:00Z","timestamp":1744588800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2025,4,14]],"date-time":"2025-04-14T00:00:00Z","timestamp":1744588800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"DOI":"10.13039\/100018700","name":"HORIZON EUROPE Climate, Energy and Mobility","doi-asserted-by":"crossref","award":["957189"],"award-info":[{"award-number":["957189"]}],"id":[{"id":"10.13039\/100018700","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Mach. Learn.: Sci. Technol."],"published-print":{"date-parts":[[2025,6,30]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    The electron density is a fundamental observable of an atomic system from which all ground-state properties can be computed. As a prediction target for machine learning (ML) models, electron density is often represented on a dense real space grid, which is data heavy, or through density fitting approximations. In this work, we show the power of targeting the density matrix (DM), a linear-scaling sparse SE(3) equivariant matrix that encodes the exact density. We introduce Graph2Mat, a universal function for converting molecular graphs into equivariant matrices. We demonstrate how a ML model that combines this Graph2Mat approach with state-of-the-art molecular graph representations can accurately predict the DM of molecular systems. The models achieve state-of-the-art performance on electron density prediction by matching the accuracy of grid-based methods, while using datasets that are at least one order of magnitude smaller. Accurately predicted electron densities can also accelerate density functional theory (DFT) calculations by reducing the number of self-consistent field (SCF) iterations needed to converge. In this work, we get an average 40% reduction on the number of SCF steps in DFT calculations of QM9 molecules with SIESTA. The novel prediction model also allows for two new and promising measures of uncertainty (total charge error and self-consistency error) that will facilitate its practical usage, e.g. within active learning workflows. These results open the door for many applications using hybrid ML-accelerated DFT methodologies, and uncertainty aware single iteration\n                    <jats:italic>ab initio<\/jats:italic>\n                    molecular dynamics.\n                  <\/jats:p>","DOI":"10.1088\/2632-2153\/adc871","type":"journal-article","created":{"date-parts":[[2025,4,2]],"date-time":"2025-04-02T18:56:28Z","timestamp":1743620188000},"page":"025013","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["Graph2Mat: universal graph to matrix conversion for electron density prediction"],"prefix":"10.1088","volume":"6","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-0904-2234","authenticated-orcid":true,"given":"Pol","family":"Febrer","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4404-7276","authenticated-orcid":false,"given":"Peter Bj\u00f8rn","family":"J\u00f8rgensen","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3621-6095","authenticated-orcid":false,"given":"Miguel","family":"Pruneda","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5138-9579","authenticated-orcid":true,"given":"Alberto","family":"Garc\u00eda","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2353-2793","authenticated-orcid":true,"given":"Pablo","family":"Ordej\u00f3n","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3198-5116","authenticated-orcid":true,"given":"Arghya","family":"Bhowmik","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"266","published-online":{"date-parts":[[2025,4,14]]},"reference":[{"key":"mlstadc871bib1","doi-asserted-by":"publisher","first-page":"170","DOI":"10.1126\/science.abn3445","article-title":"The central role of density functional theory in the AI age","volume":"381","author":"Huang","year":"2023","journal-title":"Science"},{"key":"mlstadc871bib2","doi-asserted-by":"publisher","first-page":"736","DOI":"10.1038\/s41563-021-01013-3","article-title":"Electronic-structure methods for materials design","volume":"20","author":"Marzari","year":"2021","journal-title":"Nat. Mater."},{"key":"mlstadc871bib3","doi-asserted-by":"publisher","first-page":"10142","DOI":"10.1021\/acs.chemrev.0c01111","article-title":"Machine learning force fields","volume":"121","author":"Unke","year":"2021","journal-title":"Chem. Rev."},{"key":"mlstadc871bib4","doi-asserted-by":"publisher","first-page":"750","DOI":"10.1038\/s41563-020-0777-6","article-title":"Machine-learned potentials for next-generation matter simulations","volume":"20","author":"Friederich","year":"2021","journal-title":"Nat. Mater."},{"key":"mlstadc871bib5","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/acf115","article-title":"Data efficiency and extrapolation trends in neural network interatomic potentials","volume":"4","author":"Vita","year":"2023","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstadc871bib6","doi-asserted-by":"publisher","DOI":"10.26434\/chemrxiv-2024-p5t3l","article-title":"Curator: building robust machine learning potentials for atomistic simulations autonomously with batch active learning","author":"Yang","year":"2024","journal-title":"ChemRxiv"},{"key":"mlstadc871bib7","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/ac3eb3","article-title":"Calibrated uncertainty for molecular property prediction using ensembles of message passing neural networks","volume":"3","author":"Busk","year":"2021","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstadc871bib8","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41524-022-00863-y","article-title":"Equivariant graph neural networks for fast electron density estimation of molecules, liquids and solids","volume":"8","author":"J\u00f8rgensen","year":"2022","journal-title":"npj Comput. Mater."},{"key":"mlstadc871bib9","doi-asserted-by":"publisher","first-page":"5223","DOI":"10.1038\/s41467-020-19093-1","article-title":"Quantum chemical accuracy from density functional approximations via machine learning","volume":"11","author":"Bogojeski","year":"2020","journal-title":"Nat. Commun."},{"key":"mlstadc871bib10","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/acb314","article-title":"A recipe for cracking the quantum scaling limit with machine learned electron densities","volume":"4","author":"Rackers","year":"2023","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstadc871bib11","article-title":"Image super-resolution inspired electron density prediction","author":"Li","year":"2024"},{"key":"mlstadc871bib12","doi-asserted-by":"publisher","first-page":"422","DOI":"10.1038\/s42254-021-00314-5","article-title":"Physics-informed machine learning","volume":"3","author":"Karniadakis","year":"2021","journal-title":"Nat. Rev. Phys."},{"key":"mlstadc871bib13","first-page":"pp 11423","article-title":"Mace: Higher order equivariant message passing neural networks for fast and accurate force fields","volume":"vol 35","author":"Batatia","year":"2022"},{"key":"mlstadc871bib14","doi-asserted-by":"publisher","first-page":"22","DOI":"10.1038\/s41524-019-0162-7","article-title":"Solving the electronic structure problem with machine learning","volume":"5","author":"Chandrasekaran","year":"2019","journal-title":"npj Comput. Mater."},{"key":"mlstadc871bib15","doi-asserted-by":"publisher","first-page":"161","DOI":"10.1038\/s41524-024-01343-1","article-title":"Higher-order equivariant neural networks for charge density prediction in materials","volume":"10","author":"Koker","year":"2024","journal-title":"npj Comput. Mater."},{"key":"mlstadc871bib16","doi-asserted-by":"publisher","first-page":"4451","DOI":"10.1021\/acs.jctc.2c00850","article-title":"Electronic-structure properties from atom-centered predictions of the electron density","volume":"19","author":"Grisafi","year":"2022","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc871bib17","doi-asserted-by":"publisher","first-page":"2658","DOI":"10.1021\/acs.jcim.1c00227","article-title":"Machine learning of analytical electron density in large molecules through message-passing","volume":"61","author":"Cuevas-Zuvir\u00eda","year":"2021","journal-title":"J. Chem. Inform. Modeling"},{"key":"mlstadc871bib18","doi-asserted-by":"publisher","first-page":"2745","DOI":"10.1088\/0953-8984\/14\/11\/302","article-title":"The siesta method for ab initio order-n materials simulation","volume":"14","author":"Soler","year":"2002","journal-title":"J. Phys.: Condens. Matter"},{"key":"mlstadc871bib19","doi-asserted-by":"publisher","first-page":"2175","DOI":"10.1016\/j.cpc.2009.06.022","article-title":"Ab initio molecular simulations with numeric atom-centered orbitals","volume":"180","author":"Blum","year":"2009","journal-title":"Comput. Phys. Commun."},{"key":"mlstadc871bib20","doi-asserted-by":"publisher","first-page":"2453","DOI":"10.1038\/s41467-022-29939-5","article-title":"(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials","volume":"13","author":"Batzner","year":"2022","journal-title":"Nat. Commun."},{"key":"mlstadc871bib21","first-page":"pp 14434","article-title":"Se (3)-equivariant prediction of molecular wavefunctions and electronic densities","volume":"vol 34","author":"Unke","year":"2021"},{"key":"mlstadc871bib22","doi-asserted-by":"publisher","first-page":"158","DOI":"10.1038\/s41524-022-00843-2","article-title":"Equivariant analytical mapping of first principles Hamiltonians to accurate and transferable materials models","volume":"8","author":"Zhang","year":"2022","journal-title":"npj Comput. Mater."},{"key":"mlstadc871bib23","doi-asserted-by":"publisher","first-page":"182","DOI":"10.1038\/s41524-023-01130-4","article-title":"Transferable equivariant graph neural networks for the Hamiltonians of molecules and solids","volume":"9","author":"Zhong","year":"2023","journal-title":"npj Comput. Mater."},{"key":"mlstadc871bib24","doi-asserted-by":"publisher","DOI":"10.1063\/5.0072784","article-title":"Equivariant representations for molecular Hamiltonians and n-center atomic-scale properties","volume":"156","author":"Nigam","year":"2022","journal-title":"J. Chem. Phys."},{"key":"mlstadc871bib25","doi-asserted-by":"publisher","first-page":"6281","DOI":"10.1038\/s41467-023-41953-9","article-title":"Machine learning electronic structure methods based on the one-electron reduced density matrix","volume":"14","author":"Shao","year":"2023","journal-title":"Nat. Commun."},{"key":"mlstadc871bib26","doi-asserted-by":"publisher","first-page":"4569","DOI":"10.1021\/acs.jctc.4c00042","article-title":"Predicting the one-particle density matrix with machine learning","volume":"20","author":"Hazra","year":"2024","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc871bib27","article-title":"e3nn: Euclidean neural networks","author":"Geiger","year":"2022"},{"key":"mlstadc871bib28","doi-asserted-by":"publisher","first-page":"809","DOI":"10.1002\/(SICI)1521-3951(199909)215:1<809::AID-PSSB809>3.0.CO;2-0","article-title":"Linear-scaling ab-initio calculations for large and complex systems","volume":"215","author":"Artacho","year":"1999","journal-title":"Phys. Status Solidi b"},{"key":"mlstadc871bib29","doi-asserted-by":"publisher","DOI":"10.1063\/5.0005077","article-title":"Siesta: Recent developments and applications","volume":"152","author":"Garc\u00eda","year":"2020","journal-title":"J. Chem. Phys."},{"key":"mlstadc871bib30","doi-asserted-by":"publisher","first-page":"39","DOI":"10.1016\/j.cpc.2018.01.012","article-title":"The pseudodojo: Training and grading a 85 element optimized norm-conserving pseudopotential table","volume":"226","author":"Van Setten","year":"2018","journal-title":"Comput. Phys. Commun."},{"key":"mlstadc871bib31","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.1603015","article-title":"Machine learning of accurate energy-conserving molecular force fields","volume":"3","author":"Chmiela","year":"2017","journal-title":"Sci. Adv."},{"key":"mlstadc871bib32","doi-asserted-by":"publisher","DOI":"10.1038\/sdata.2014.22","article-title":"Quantum chemistry structures and properties of 134 kilo molecules","volume":"1","author":"Ramakrishnan","year":"2014","journal-title":"Sci. Data"},{"key":"mlstadc871bib33","doi-asserted-by":"publisher","author":"J\u00f8rgensen","year":"2022","DOI":"10.11583\/DTU.16691825)"},{"key":"mlstadc871bib34","doi-asserted-by":"publisher","DOI":"10.26434\/chemrxiv-2024-j4g21)","article-title":"Density matrix, Hamiltonian matrix, overlap matrix and energy density matrix of qm9 and ethylene carbonate dataset","author":"Bhowmik","year":"2024"},{"key":"mlstadc871bib35","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.64.235111","article-title":"Numerical atomic orbitals for linear-scaling calculations","volume":"64","author":"Junquera","year":"2001","journal-title":"Phys. Rev. B"},{"key":"mlstadc871bib36","first-page":"pp 7865","article-title":"On over-squashing in message passing neural networks: the impact of width, depth and topology","author":"Di Giovanni","year":"2023"},{"key":"mlstadc871bib37","doi-asserted-by":"publisher","first-page":"579","DOI":"10.1038\/s41467-023-36329-y","article-title":"Learning local equivariant representations for large-scale atomistic dynamics","volume":"14","author":"Musaelian","year":"2023","journal-title":"Nat. Commun."},{"key":"mlstadc871bib38","first-page":"pp 9377","article-title":"Equivariant message passing for the prediction of tensorial properties and molecular spectra","author":"Sch\u00fctt","year":"2021"},{"key":"mlstadc871bib39","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-023-46382-8","article-title":"A universal framework for accurate and efficient geometric deep learning of molecular systems","volume":"13","author":"Zhang","year":"2023","journal-title":"Sci. Rep."},{"key":"mlstadc871bib40","doi-asserted-by":"publisher","first-page":"1833","DOI":"10.1063\/1.1740588","article-title":"Electronic population analysis on LCAO\u2013MO molecular wave functions. I","volume":"23","author":"Mulliken","year":"1955","journal-title":"J. Chem. Phys."},{"key":"mlstadc871bib41","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.597181)","article-title":"sisl: v0.14.3","author":"Papior","year":"2024"},{"key":"mlstadc871bib42","doi-asserted-by":"publisher","first-page":"4771","DOI":"10.1103\/PhysRevB.47.4771","article-title":"Accelerating the convergence of force calculations in electronic-structure computations","volume":"47","author":"Chan","year":"1993","journal-title":"Phys. Rev. B"},{"key":"mlstadc871bib43","doi-asserted-by":"publisher","first-page":"712","DOI":"10.1093\/jxb\/erp043","article-title":"Known knowns, known unknowns, unknown unknowns and the propagation of scientific enquiry","volume":"60","author":"Logan","year":"2009","journal-title":"J. Exp. Bot."}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,4,14]],"date-time":"2025-04-14T06:27:43Z","timestamp":1744612063000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc871"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4,14]]},"references-count":43,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2025,4,14]]},"published-print":{"date-parts":[[2025,6,30]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/adc871","relation":{"has-preprint":[{"id-type":"doi","id":"10.26434\/chemrxiv-2024-j4g21","asserted-by":"object"},{"id-type":"doi","id":"10.26434\/chemrxiv-2024-j4g21-v2","asserted-by":"object"}]},"ISSN":["2632-2153"],"issn-type":[{"value":"2632-2153","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4,14]]},"assertion":[{"value":"Graph2Mat: universal graph to matrix conversion for electron density prediction","name":"article_title","label":"Article Title"},{"value":"Machine Learning: Science and Technology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2025 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2024-08-08","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-04-01","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-04-14","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}