<link rel="stylesheet" href="/res/styles.39be4579664c3a48.css">
Artículo

Advancing Multiscale Molecular Modeling with Machine Learning-Derived Electrostatics

Semelak, Jonathan Alexis Icon; Pickering, Ignacio ; Huddleston, Kate ; Olmos, Justo Facundo Icon; Grassano, Juan Santiago ; Clemente, Camila Mara Icon; Drusin, Salvador Iván Icon; Marti, Marcelo Adrian Icon; González Lebrero, Mariano Camilo Icon; Roitberg, Adrian Icon; Estrin, Dario Ariel Icon

Otros responsables

Fecha

03/2025

Editorial

American Chemical Society

Revista

Journal of Chemical Theory and Computation

Título del libro

Nombre del evento

Fecha del evento

Institución organizadora

ISSN

1549-9618

eISSN

ISBN

Idioma

Inglés

Tipo del evento

Clasificación temática

Resumen

We introduce an innovative machine learning (ML)-basedframework for multiscale molecular modeling in which the ML subsystem istreated as an electrostatic entity interacting with its molecular mechanics (MM)environment through classical electrostatics. The integration of ML accuracy withmultiscale modeling is accomplished by leveraging the capabilities of the ANIneural networks to predict geometry-dependent atomic partial charges at theminimal basis iterative stockholder (MBIS) level, going beyond static mechanicalembedding. This ML/MM approach can closely approximate state-of-the-artmultiscale quantum-classical (QM/MM) methods while significantly loweringcomputational requirements, thereby facilitating more efficient and precisesimulations in computational chemistry. The method requires no additionaltraining beyond the initial model setup and is integrated into Amber, one of themost widely used software suites for molecular modeling, ensuring accessibility tothe broader community. We validate its performance across a variety of challenging applications, including the solvation structure,vibrational spectra, torsion free energy profiles, and protein−ligand interactions, achieving excellent agreement with QM/MMbenchmarks. This framework not only advances the frontiers of multiscale modeling but also showcases the potential of machinelearning to achieve quantum-level accuracy with exceptional efficiency for complex chemical systems.

Aprobación

Revisión

Complementado por

Referenciado por

Licencia Creative Commons

Excepto donde se indique lo contrario, la licencia de este ítem se describe como by-nc-sa