Realistic phase diagram of water from “first principles” data-driven quantum simulations

Nature Communications, Journal Year: 2023, Volume and Issue: 14(1)

Published: June 8, 2023

Since the experimental characterization of low-pressure region water's phase diagram in early 1900s, scientists have been on a quest to understand thermodynamic stability ice polymorphs molecular level. In this study, we demonstrate that combining MB-pol data-driven many-body potential for water, which was rigorously derived from "first principles" and exhibits chemical accuracy, with advanced enhanced-sampling algorithms, correctly describe quantum nature motion equilibria, enables computer simulations an unprecedented level realism. Besides providing fundamental insights into how enthalpic, entropic, nuclear effects shape free-energy landscape recent progress simulations, encode interactions, has opened door realistic computational studies complex systems, bridging gap between experiments simulations.

Language: Английский

---

MBX: A many-body energy and force calculator for data-driven many-body simulations

The Journal of Chemical Physics, Journal Year: 2023, Volume and Issue: 159(5)

Published: Aug. 1, 2023

Many-Body eXpansion (MBX) is a C++ library that implements many-body potential energy functions (PEFs) within the "many-body energy" (MB-nrg) formalism. MB-nrg PEFs integrate an underlying polarizable model with explicit machine-learned representations of interactions to achieve chemical accuracy from gas condensed phases. MBX can be employed either as stand-alone package or energy/force engine integrated generic software for molecular dynamics and Monte Carlo simulations. parallelized internally using Open Multi-Processing utilize Message Passing Interface when available in interfaced simulation software. enables classical quantum simulations PEFs, well hybrid combine conventional force fields diverse systems ranging small gas-phase clusters aqueous solutions fluids biomolecular metal-organic frameworks.

Language: Английский

---

Enhancing the Quality and Reliability of Machine Learning Interatomic Potentials through Better Reporting Practices

The Journal of Physical Chemistry C, Journal Year: 2024, Volume and Issue: 128(16), P. 6524 - 6537

Published: March 20, 2024

Recent developments in machine learning interatomic potentials (MLIPs) have empowered even nonexperts to train MLIPs for accelerating materials simulations. However, reproducibility and independent evaluation of presented MLIP results is hindered by a lack clear standards current literature. In this Perspective, we aim provide guidance on best practices documenting use while walking the reader through development deployment including hardware software requirements, generating training data, models, validating predictions, inference. We also suggest useful plotting analyses validate boost confidence deployed models. Finally, step-by-step checklist practitioners directly before publication standardize information be reported. Overall, hope that our work will encourage reliable reproducible these MLIPs, which accelerate their ability make positive impact various disciplines science, chemistry, biology, among others.

Language: Английский

---

Many-body interactions and deep neural network potentials for water

The Journal of Chemical Physics, Journal Year: 2024, Volume and Issue: 160(14)

Published: April 8, 2024

We present a detailed assessment of deep neural network potentials developed within the Deep Potential Molecular Dynamics (DeePMD) framework and trained on MB-pol data-driven many-body potential energy function. Specific focus is directed at ability DeePMD-based to correctly reproduce accuracy across various water systems. Analyses bulk interfacial properties as well interactions characteristic elucidate inherent limitations in transferability predictive potentials. These can be traced back an incomplete implementation "nearsightedness electronic matter" principle, which may common throughout machine learning that do not include proper representation self-consistently determined long-range electric fields. findings provide further support for "short-blanket dilemma" faced by potentials, highlighting challenges achieving balance between computational efficiency rigorous, physics-based water. Finally, we believe our study contributes ongoing discourse development application models simulating systems, offering insights could guide future improvements field.

Language: Английский

---

No Headache for PIPs: A PIP Potential for Aspirin Runs Much Faster and with Similar Precision Than Other Machine-Learned Potentials

Journal of Chemical Theory and Computation, Journal Year: 2024, Volume and Issue: 20(8), P. 3008 - 3018

Published: April 9, 2024

Assessments of machine-learning (ML) potentials are an important aspect the rapid development this field. We recently reported assessment linear-regression permutationally invariant polynomial (PIP) method for ethanol, using widely used (revised) rMD17 data set. demonstrated that PIP approach outperformed numerous other methods, e.g., ANI, PhysNet, sGDML, and p-KRR, with respect to precision notably speed [Houston et al., J. Chem. Phys. 2022, 156, 044120]. Here, we extend 21-atom aspirin molecule, set, a focus on evaluation. Both energies forces training, several PIPs is examined both. Normal mode frequencies, methyl torsional potential, 1d vibrational OH stretch presented. show achieves level obtained from ML atom-centered neural network linear regression ACE, kernel as by Kovács al. in Theory Comput. 2021, 17, 7696–7711. More significantly, PESs run much faster than all whose timings were evaluated paper. also PES extrapolates well enough describe internal motions aspirin, including stretch.

Language: Английский

---

Targeted Transferable Machine-Learned Potential for Linear Alkanes Trained on C₁₄H₃₀ and Tested for C₄H₁₀ to C₃₀H₆₂

Journal of Chemical Theory and Computation, Journal Year: 2025, Volume and Issue: unknown

Published: March 27, 2025

Given the great importance of linear alkanes in fundamental and applied research, an accurate machine-learned potential (MLP) would be a major advance computational modeling these hydrocarbons. Recently, we reported novel, many-body permutationally invariant model that was trained specifically for 44-atom hydrocarbon C14H30 on roughly 250,000 B3LYP energies (Qu, C.; Houston, P. L.; Allison, T.; Schneider, B. I.; Bowman, J. M. Chem. Theory Comput. 2024, 20, 9339–9353). Here, demonstrate accuracy transferability this ranging from butane C4H10 up to C30H62. Unlike other approaches aim universal applicability, present approach is targeted alkanes. The mean absolute error (MAE) energy ranges 0.26 kcal/mol rises 0.73 C30H62 over range 80 600 These values are unprecedented transferable potentials indicate high performance potential. conformational barriers shown excellent agreement with high-level ab initio calculations pentane, largest alkane which such have been reported. Vibrational power spectra molecular dynamics presented briefly discussed. Finally, evaluation time vary linearly number atoms.

Language: Английский

---

Data-driven many-body potentials from density functional theory for aqueous phase chemistry

Chemical Physics Reviews, Journal Year: 2023, Volume and Issue: 4(1)

Published: Jan. 10, 2023

Density functional theory (DFT) has been applied to modeling molecular interactions in water for over three decades. The ubiquity of chemical and biological processes demands a unified understanding its physics, from the single molecule thermodynamic limit everything between. Recent advances development data-driven machine-learning potentials have accelerated simulation aqueous systems with DFT accuracy. However, anomalous properties condensed phase, where rigorous treatment both local non-local many-body (MB) is order, are often unsatisfactory or partially missing models water. In this review, we discuss based on provide comprehensive description general theoretical/computational framework reference data. This framework, coined MB-DFT, readily enables efficient dynamics (MD) simulations small molecules, gas phases, while preserving accuracy underlying model. Theoretical considerations emphasized, including role that delocalization error plays MB-DFT possibility elevate near-chemical-accuracy through density-corrected formalism. described detail, along application MB-MD recent extension reactive solution within quantum mechanics/MB mechanics (QM/MB-MM) scheme, using as prototypical solvent. Finally, identify open challenges future directions QM/MB-MM phases.

Language: Английский

---

Molecular Insights into the Influence of Ions on the Water Structure. I. Alkali Metal Ions in Solution

The Journal of Physical Chemistry B, Journal Year: 2024, Volume and Issue: 128(8), P. 1953 - 1962

Published: Feb. 19, 2024

In this study, we explore the impact of alkali metal ions (Li+, Na+, K+, Rb+, and Cs+) on hydration structure water using molecular dynamics simulations carried out with MB-nrg potential energy functions (PEFs). Our analyses include radial distribution functions, coordination numbers, dipole moments, infrared spectra molecules, calculated as a function solvation shells. The results collectively indicate highly local influence all hydrogen-bond network established by surrounding smallest most densely charged Li+ ion exerting pronounced effect. Remarkably, PEFs demonstrate excellent agreement available experimental data for position size first shells, underscoring their predictive models realistic ionic aqueous solutions across various thermodynamic conditions environments.

Language: Английский

---

Current Status of the MB-pol Data-Driven Many-Body Potential for Predictive Simulations of Water Across Different Phases

Journal of Chemical Theory and Computation, Journal Year: 2024, Volume and Issue: 20(21), P. 9269 - 9289

Published: Oct. 14, 2024

Developing a molecular-level understanding of the properties water is central to numerous scientific and technological applications. However, accurately modeling through computer simulations has been significant challenge due complex nature hydrogen-bonding network that molecules form under different thermodynamic conditions. This complexity led over five decades research many attempts. The introduction MB-pol data-driven many-body potential energy function marked advancement toward universal molecular model capable predicting structural, thermodynamic, dynamical, spectroscopic across all phases. By integrating physics-based (i.e., machine-learned) components, which correctly capture delicate balance among interactions, achieves chemical accuracy, enabling realistic water, from gas-phase clusters liquid ice. In this review, we present comprehensive overview formalism adopted by MB-pol, highlight main results predictions made with date, discuss prospects for future extensions potentials generic reactive systems.

Language: Английский

---

DFT-Based Permutationally Invariant Polynomial Potentials Capture the Twists and Turns of C₁₄H₃₀

Journal of Chemical Theory and Computation, Journal Year: 2024, Volume and Issue: 20(21), P. 9339 - 9353

Published: Oct. 21, 2024

Hydrocarbons are ubiquitous as fuels, solvents, lubricants, and the principal components of plastics fibers, yet our ability to predict their dynamical properties is limited force-field mechanics. Here, we report two machine-learned potential energy surfaces (PESs) for linear 44-atom hydrocarbon C

Language: Английский

---