Physics-inspired equivariant descriptors of non-bonded interactions
JSON Export
{
"id": "1924",
"updated": "2023-10-03T08:50:09.033577+00:00",
"metadata": {
"version": 1,
"contributors": [
{
"givennames": "Kevin K.",
"affiliations": [
"Laboratory of Computational Science and Modelling (COSMO), \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Vaud, Switzerland"
],
"familyname": "Huguenin-Dumittan"
},
{
"givennames": "Philip",
"affiliations": [
"Laboratory of Computational Science and Modelling (COSMO), \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Vaud, Switzerland"
],
"familyname": "Loche"
},
{
"givennames": "Haoran",
"affiliations": [
"Laboratory of Computational Science and Modelling (COSMO), \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Vaud, Switzerland"
],
"familyname": "Ni"
},
{
"givennames": "Michele",
"affiliations": [
"Laboratory of Computational Science and Modelling (COSMO), \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Vaud, Switzerland"
],
"email": "michele.ceriotti@epfl.ch",
"familyname": "Ceriotti"
}
],
"title": "Physics-inspired equivariant descriptors of non-bonded interactions",
"_oai": {
"id": "oai:materialscloud.org:1924"
},
"keywords": [
"ERC",
"machine learning",
"long-range interactions",
"electrostatics",
"dispersion"
],
"publication_date": "Oct 03, 2023, 10:50:08",
"_files": [
{
"key": "point_charges_Training_set_p1.xyz",
"description": "Dataset containing the structures and energies of \"NaCl\" atoms interacting via a pure Coulomb potential",
"checksum": "md5:9e76b4a8972a2e829335163fe31c7d7c",
"size": 7154931
},
{
"key": "point_charges_Training_set_p6.xyz",
"description": "Dataset containing the structures and energies of atoms interacting via a pure dispersion (1/r^6) potential",
"checksum": "md5:57c54e0ea0b01be89e0a0622f0f2ff96",
"size": 7155390
},
{
"key": "bio_dimers.xyz",
"description": "Dataset containing structures and DFT energies & forces of dimers taken from sidechain-sidechain fragments in biomolecules",
"checksum": "md5:cdf5f3e139bd56a2c31014c8e50e8b93",
"size": 69468453
},
{
"key": "bio_dimers_monomers.xyz",
"description": "The structures and DFT energies & forces of the individual monomers contained in the \"bio_dimers.xyz\" dataset",
"checksum": "md5:3d4b87ec5e1a79b75c41c48a78f78552",
"size": 5113854
},
{
"key": "bio_dimers_control.in",
"description": "The FHI-AIMS input file to compute the energies of the \"bio_dimers.xyz\" and \"bio_dimers_monomers.xyz\" structures",
"checksum": "md5:47e973400e61374c26fb2cedbf12838d",
"size": 17109
},
{
"key": "xenon.xyz",
"description": "Dataset containing structures and DFT energies & forces of Xenon dimers and trimers",
"checksum": "md5:9371ed17727e39635fa0c11e27425ace",
"size": 41908
},
{
"key": "xenon_control.in",
"description": "The FHI-AIMS input file to compute the energies and forces of the structures in \"xenon.xyz\"",
"checksum": "md5:4ddc62a1c4d83bf68518010daf5903bd",
"size": 2685
}
],
"references": [
{
"comment": "Preprint in which the data is discussed",
"citation": "Kevin K. Huguenin-Dumittan, Philip Loche, Ni Haoran, Michele Ceriotti, Arxiv (submitted 2023).",
"url": "https://arxiv.org/abs/2308.13208",
"type": "Preprint"
},
{
"comment": "Paper in which some of the datasets were originally introduced",
"doi": "10.1063/1.5128375",
"citation": "Andrea Grisafi, Michele Ceriotti, J. Chem. Phys. 151, 204105 (2019)",
"url": "https://doi.org/10.1063/1.5128375",
"type": "Journal reference"
}
],
"description": "One essential ingredient in many machine learning (ML) based methods for atomistic modeling of materials and molecules is the use of locality. While allowing better system-size scaling, this systematically neglects long-range (LR) effects, such as electrostatics or dispersion interaction. We present an extension of the long distance equivariant (LODE) framework that can handle diverse LR interactions in a consistent way, and seamlessly integrates with preexisting methods by building new sets of atom centered features. We provide a direct physical interpretation of these using the multipole expansion, which allows for simpler and more efficient implementations. The framework is applied to simple toy systems as proof of concept, and a heterogeneous set of molecular dimers to push the method to its limits. By generalizing LODE to arbitrary asymptotic behaviors, we provide a coherent approach to treat arbitrary two- and many-body non-bonded interactions in the data-driven modeling of matter.",
"status": "published",
"license": "Creative Commons Attribution 4.0 International",
"conceptrecid": "1923",
"is_last": true,
"mcid": "2023.151",
"edited_by": 576,
"id": "1924",
"owner": 1147,
"license_addendum": null,
"doi": "10.24435/materialscloud:23-99"
},
"revision": 6,
"created": "2023-10-02T08:36:05.743665+00:00"
}