A machine learning model of chemical shifts for chemically and structurally diverse molecular solids
JSON Export
{
"created": "2022-10-28T08:49:06.070494+00:00",
"metadata": {
"references": [
{
"citation": "M. Cordova, E. A. Engel, A. Stefaniuk, F. Paruzzo, A. Hofstetter, M. Ceriotti, L. Emsley, J. Phys. Chem. C 126, 16710-16720 (2022)",
"url": "https://pubs.acs.org/doi/full/10.1021/acs.jpcc.2c03854",
"comment": "Paper in which the method is described",
"doi": "10.1021/acs.jpcc.2c03854",
"type": "Journal reference"
}
],
"mcid": "2022.147",
"id": "1503",
"is_last": true,
"title": "A machine learning model of chemical shifts for chemically and structurally diverse molecular solids",
"publication_date": "Nov 11, 2022, 18:24:48",
"edited_by": 576,
"_oai": {
"id": "oai:materialscloud.org:1503"
},
"version": 1,
"description": "Nuclear magnetic resonance (NMR) chemical shifts are a direct probe of local atomic environments and can be used to determine the structure of solid materials. However, the substantial computational cost required to predict accurate chemical shifts is a key bottleneck for NMR crystallography. We recently introduced ShiftML, a machine-learning model of chemical shifts in molecular solids, trained on minimum-energy geometries of materials composed of C, H, N, O, and S that provides rapid chemical shift predictions with density functional theory (DFT) accuracy. Here, we extend the capabilities of ShiftML to predict chemical shifts for both finite temperature structures and more chemically diverse compounds, while retaining the same speed and accuracy. For a benchmark set of 13 molecular solids, we find a root-mean-squared error of 0.47 ppm with respect to experiment for 1H shift predictions (compared to 0.35 ppm for explicit DFT calculations), while reducing the computational cost by over four orders of magnitude.",
"status": "published",
"license_addendum": null,
"keywords": [
"machine learning",
"MARVEL/DD1",
"SNSF"
],
"license": "Creative Commons Attribution 4.0 International",
"owner": 339,
"contributors": [
{
"affiliations": [
"Institut des Sciences et Ing\u00e9nierie Chimiques, \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland"
],
"familyname": "Cordova",
"email": "manuel.cordova@epfl.ch",
"givennames": "Manuel"
},
{
"affiliations": [
"Theory of Condensed Matter Group, Cavendish Laboratory, University of Cambridge, J. J. Thomson Avenue, Cambridge CB3 0HE, U.K."
],
"familyname": "Engel",
"givennames": "Edgar A."
},
{
"affiliations": [
"Institut des Sciences et Ing\u00e9nierie Chimiques, \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland"
],
"familyname": "Stefaniuk",
"givennames": "Artur"
},
{
"affiliations": [
"Institut des Sciences et Ing\u00e9nierie Chimiques, \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland"
],
"familyname": "Paruzzo",
"givennames": "Federico"
},
{
"affiliations": [
"Institut des Sciences et Ing\u00e9nierie Chimiques, \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland"
],
"familyname": "Hofstetter",
"givennames": "Albert"
},
{
"affiliations": [
"Institut des mat\u00e9riaux, \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland"
],
"familyname": "Ceriotti",
"email": "michele.ceriotti@epfl.ch",
"givennames": "Michele"
},
{
"affiliations": [
"Institut des Sciences et Ing\u00e9nierie Chimiques, \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland"
],
"familyname": "Emsley",
"email": "lyndon.emsley@epfl.ch",
"givennames": "Lyndon"
}
],
"conceptrecid": "1502",
"doi": "10.24435/materialscloud:a9-4n",
"_files": [
{
"size": 56634,
"key": "ShiftML2.zip",
"description": "ShiftML2 Python package",
"checksum": "md5:12d0872eb54908dc6ea6244895b4da0e"
},
{
"size": 472732,
"key": "Experimental_benchmark_nmr.zip",
"description": "GIPAW NMR computations performed on a set of crystal structures used to compare ShiftML2 with experimental chemical shifts",
"checksum": "md5:68520acd7410ffdd60627679501b2c38"
},
{
"size": 120941801,
"key": "final_training_test_sets.zip",
"description": "Final set of training and test environments for each nucleus",
"checksum": "md5:7fda429b5eadf7378ec1903ca19cd278"
},
{
"size": 40844081,
"key": "initial_training_test_sets.zip",
"description": "Initial set of training and test environments for each nucleus",
"checksum": "md5:5bc1ddc62a1fb532e8707b723a5de6ab"
},
{
"size": 6578532,
"key": "relax_vs_md.zip",
"description": "Snapshots from AIMD simulation used to compare the accuracy of the model in relaxed and distorted structures",
"checksum": "md5:a4c9f521cf6352704e9fd477b7767967"
},
{
"size": 550092,
"key": "ShiftML_poly.zip",
"description": "Set of candidate structures with GIPAW NMR computations for three molecular solids",
"checksum": "md5:28b1c98148fef74cd3991adf97f3f55e"
},
{
"size": 2888914,
"key": "training_scripts.zip",
"description": "Python notebook scripts used to train the model",
"checksum": "md5:2a952934a684939ef148beb13116a1da"
},
{
"size": 3758893985,
"key": "ShiftML_v2_model_H.pk",
"description": "ShiftML model (kernel) for hydrogen",
"checksum": "md5:0155e4f88b00a52c7ee7555012961149"
},
{
"size": 3552070952,
"key": "ShiftML_v2_model_C.pk",
"description": "ShiftML model (kernel) for carbon",
"checksum": "md5:915ddef4e10c74b1d92215315500a289"
},
{
"size": 3670667122,
"key": "ShiftML_v2_model_N.pk",
"description": "ShiftML model (kernel) for nitrogen",
"checksum": "md5:2f0fce6d788c6b7c07d917454dcff6a5"
},
{
"size": 3777679999,
"key": "ShiftML_v2_model_O.pk",
"description": "ShiftML model (kernel) for oxygen",
"checksum": "md5:421a80c90cdd54110dcbe96b94cb3ab5"
},
{
"size": 1565464921,
"key": "ShiftML_v2_model_F.pk",
"description": "ShiftML model (kernel) for fluorine",
"checksum": "md5:71a8c7dc1b010f76857e2cc41bcf2e16"
},
{
"size": 1104021149,
"key": "ShiftML_v2_model_S.pk",
"description": "ShiftML model (kernel) for sulfur",
"checksum": "md5:8373a1dcc7065aec7dbc91d2f6cb17ab"
},
{
"size": 380021978,
"key": "ShiftML_v2_model_P.pk",
"description": "ShiftML model (kernel) for phosphorus",
"checksum": "md5:6f1c6a6cdbb60de78ddb9cac0edf3ec0"
},
{
"size": 989833481,
"key": "ShiftML_v2_model_Cl.pk",
"description": "ShiftML model (kernel) for chlorine",
"checksum": "md5:d92aefad57270115a756fd3e301a8de5"
},
{
"size": 54211646,
"key": "ShiftML_v2_model_Na.pk",
"description": "ShiftML model (kernel) for sodium",
"checksum": "md5:7fd44392f47b46730e22d1adf5576dd7"
},
{
"size": 26497099,
"key": "ShiftML_v2_model_Ca.pk",
"description": "ShiftML model (kernel) for calcium",
"checksum": "md5:0be6d72103c591f8239164f922fbbad4"
},
{
"size": 46686067,
"key": "ShiftML_v2_model_K.pk",
"description": "ShiftML model (kernel) for potassium",
"checksum": "md5:1d9e1ae13b2e8c97b82ff16cc216bfe3"
},
{
"size": 12002530,
"key": "ShiftML_v2_model_Mg.pk",
"description": "ShiftML model (kernel) for magnesium",
"checksum": "md5:407b0ee50683d1296e6d9480413cc165"
},
{
"size": 1274,
"key": "README.txt",
"description": "File descriptions and instructions for package installation",
"checksum": "md5:c4bcccd488166a2df257f652412574aa"
}
]
},
"id": "1503",
"updated": "2022-11-11T17:24:48.966777+00:00",
"revision": 6
}