MLatom 3: A Platform for Machine Learning-Enhanced Computational Chemistry Simulations and Workflows

Journal of Chemical Theory and Computation, Journal Year: 2024, Volume and Issue: 20(3), P. 1193 - 1213

Published: Jan. 25, 2024

Machine learning (ML) is increasingly becoming a common tool in computational chemistry. At the same time, rapid development of ML methods requires flexible software framework for designing custom workflows. MLatom 3 program package designed to leverage power enhance typical chemistry simulations and create complex This open-source provides plenty choice users who can run with command-line options, input files, or scripts using as Python package, both on their computers online XACS cloud computing service at XACScloud.com. Computational chemists calculate energies thermochemical properties, optimize geometries, molecular quantum dynamics, simulate (ro)vibrational, one-photon UV/vis absorption, two-photon absorption spectra ML, mechanical, combined models. The choose from an extensive library containing pretrained models mechanical approximations such AIQM1 approaching coupled-cluster accuracy. developers build own various algorithms. great flexibility largely due use interfaces many state-of-the-art packages libraries.

Language: Английский

---

ANI-1ccx-gelu Universal Interatomic Potential and Its Fine-Tuning: Toward Accurate and Efficient Anharmonic Vibrational Frequencies

The Journal of Physical Chemistry Letters, Journal Year: 2025, Volume and Issue: unknown, P. 483 - 493

Published: Jan. 2, 2025

Calculating anharmonic vibrational modes of molecules for interpreting experimental spectra is one the most interesting challenges contemporary computational chemistry. However, traditional QM methods are costly this application. Machine learning techniques have emerged as a powerful tool substituting methods. Universal interatomic potentials (UIPs) hold particular promise to deliver accurate results at fraction cost methods, but performance UIPs calculating frequencies remains hitherto unknown. Here we show that despite known excellent representative UIP ANI-1ccx thermochemical properties, it fails due original unfortunate choice activation function. Hence, recommend evaluating new on an additional important quality test. To remedy shortcomings ANI-1ccx, introduce its reformulation ANI-1ccx-gelu with GELU function, which capable IR reasonable accuracy (close B3LYP/6-31G*). We also our can be fine-tuned obtain very some specific more effort needed improve overall and capability fine-tuning. The will included part universal updatable AI-enhanced (UAIQM) platform available together usage fine-tuning tutorials in open-source MLatom https://github.com/dralgroup/mlatom. calculations performed via web browser https://XACScloud.com.

Language: Английский

---

Informing geometric deep learning with electronic interactions to accelerate quantum chemistry

Proceedings of the National Academy of Sciences, Journal Year: 2022, Volume and Issue: 119(31)

Published: July 28, 2022

Predicting electronic energies, densities, and related chemical properties can facilitate the discovery of novel catalysts, medicines, battery materials. However, existing machine learning techniques are challenged by scarcity training data when exploring unknown spaces. We overcome this barrier systematically incorporating knowledge molecular structure into deep learning. By developing a physics-inspired equivariant neural network, we introduce method to learn representations based on interactions among atomic orbitals. Our method, OrbNet-Equi, leverages efficient tight-binding simulations learned mappings recover high-fidelity physical quantities. OrbNet-Equi accurately models wide spectrum target while being several orders magnitude faster than density functional theory. Despite only using samples collected from readily available small-molecule libraries, outperforms traditional semiempirical learning-based methods comprehensive downstream benchmarks that encompass diverse main-group processes. also describes in challenging charge-transfer complexes open-shell systems. anticipate strategy presented here will help expand opportunities for studies chemistry materials science, where acquisition experimental or reference is costly.

Language: Английский

---

Δ-Quantum machine-learning for medicinal chemistry

Physical Chemistry Chemical Physics, Journal Year: 2022, Volume and Issue: 24(18), P. 10775 - 10783

Published: Jan. 1, 2022

Many molecular design tasks benefit from fast and accurate calculations of quantum-mechanical (QM) properties. However, the computational cost QM methods applied to drug-like molecules currently renders large-scale applications quantum chemistry challenging. Aiming mitigate this problem, we developed DelFTa, an open-source toolbox for prediction electronic properties at density functional (DFT) level theory, using Δ-machine-learning. Δ-Learning corrects error (Δ) a but inaccurate property calculation. DelFTa employs state-of-the-art three-dimensional message-passing neural networks trained on large dataset It provides access wide array observables molecular, atomic bond levels by predicting approximations DFT values low-cost semiempirical baseline. outperformed its direct-learning counterpart most considered endpoints. The results suggest that predictions non-covalent intra- intermolecular interactions can be extrapolated larger biomolecular systems. software is fully open-sourced features documented command-line Python APIs.

Language: Английский

---

Modern Alchemical Free Energy Methods for Drug Discovery Explained

ACS Physical Chemistry Au, Journal Year: 2023, Volume and Issue: 3(6), P. 478 - 491

Published: Oct. 4, 2023

This Perspective provides a contextual explanation of the current state-of-the-art alchemical free energy methods and their role in drug discovery as well highlights select emerging technologies. The narrative attempts to answer basic questions about what goes on "under hood" simulations provide general guidelines for how run analyze results. It is hope that this work will valuable introduction students scientists field.

Language: Английский

---

UniversalQM/MMapproaches for general nanoscale applications

Wiley Interdisciplinary Reviews Computational Molecular Science, Journal Year: 2023, Volume and Issue: 13(4)

Published: Feb. 1, 2023

Abstract Quantum mechanics/molecular mechanics (QM/MM) hybrid models allow one to address chemical phenomena in complex molecular environments. Whereas this modeling approach can cope with a large system size at moderate computational costs, the are often tedious construct and require manual preprocessing expertise. As result, transferability new application areas be limited many parameters not easy adjust reference data that typically scarce. Therefore, it is desirable devise automated procedures of controllable accuracy, which enables such standardized black‐box‐type manner. Although diverse best‐practice protocols have been set up for construction individual components QM/MM model (e.g., MM potential, type embedding, choice QM region), reconcile all steps still rare. Here, we review state art focus on automation. We elaborate parametrization, atom‐economical physically‐motivated region selection, embedding schemes incorporate mutual polarization as critical model. In view broad scope field, mostly restrict discussion methodologies build de novo based first‐principles data, uncertainty quantification, error mitigation high potential Ultimately, able reliable fast efficient way without being constrained by specific or technical limitations. This article categorized under: Electronic Structure Theory > Combined Methods

Language: Английский

---

AIMNet2: A Neural Network Potential to Meet your Neutral, Charged, Organic, and Elemental-Organic Needs

Published: Oct. 12, 2023

Machine learned interatomic potentials (MLIPs) are reshaping computational chemistry practices because of their ability to drastically exceed the accuracy-length/time scale tradeoff. Despite this attraction, benefits such efficiency only impactful when an MLIP uniquely enables insight into a target system or is broadly transferable outside training dataset, where models achieving latter seldom reported. In work, we present 2nd generation our atoms-in-molecules neural network potential (AIMNet2), which applicable species composed up 14 chemical elements in both neutral and charged states, making it valuable model for modeling majority non-metallic compounds. Using exhaustive dataset 20 million hybrid quantum calculations, AIMNet2 combines ML-parameterized short-range physics-based long-range terms attain generalizability that reaches from simple organics diverse molecules with “exotic” element-organic bonding. We show outperforms semi-empirical GFN-xTB on par reference density functional theory interaction energy contributions, conformer search tasks, torsion rotation profiles, molecular-to-macromolecular geometry optimization. Overall, demonstrated coverage significant step toward providing access MLIPs avoid crucial limitation curating additional data retraining each new application.

Language: Английский

---

QDπ: A Quantum Deep Potential Interaction Model for Drug Discovery

Journal of Chemical Theory and Computation, Journal Year: 2023, Volume and Issue: 19(4), P. 1261 - 1275

Published: Jan. 25, 2023

We report QDπ-v1.0 for modeling the internal energy of drug molecules containing H, C, N, and O atoms. The QDπ model is in form a quantum mechanical/machine learning potential correction (QM/Δ-MLP) that uses fast third-order self-consistent density-functional tight-binding (DFTB3/3OB) corrected to quantitatively high-level accuracy through deep-learning (DeepPot-SE). has advantage it able properly treat electrostatic interactions handle changes charge/protonation states. trained against reference data computed at ωB97X/6-31G* level (as ANI-1x set) compared several other approximate semiempirical machine potentials (ANI-1x, ANI-2x, DFTB3, MNDO/d, AM1, PM6, GFN1-xTB, GFN2-xTB). demonstrated be accurate wide range intra- intermolecular (despite its intended use as an model) shown perform exceptionally well relative protonation/deprotonation energies tautomers. An example application reactions involved RNA strand cleavage catalyzed by protein nucleic acid enzymes illustrates average errors less than 0.5 kcal/mol, whereas models have over order magnitude greater. Taken together, this makes highly attractive force field discovery.

Language: Английский

---

Transfer learning for chemically accurate interatomic neural network potentials

Physical Chemistry Chemical Physics, Journal Year: 2023, Volume and Issue: 25(7), P. 5383 - 5396

Published: Jan. 1, 2023

Developing machine learning-based interatomic potentials from ab initio electronic structure methods remains a challenging task for computational chemistry and materials science. This work studies the capability of transfer learning, in particular discriminative fine-tuning, efficiently generating chemically accurate neural network on organic molecules MD17 ANI data sets. We show that pre-training parameters obtained density functional calculations considerably improves sample efficiency models trained more data. Additionally, we fine-tuning with energy labels alone can suffice to obtain atomic forces run large-scale atomistic simulations, provided well-designed set. also investigate possible limitations especially regarding design size Finally, provide GM-NN pre-trained fine-tuned ANI-1x ANI-1ccx sets, which easily be applied molecules.

Language: Английский

---

In Silico Chemical Experiments in the Age of AI: From Quantum Chemistry to Machine Learning and Back

Advanced Materials, Journal Year: 2024, Volume and Issue: 36(30)

Published: May 25, 2024

Abstract Computational chemistry is an indispensable tool for understanding molecules and predicting chemical properties. However, traditional computational methods face significant challenges due to the difficulty of solving Schrödinger equations increasing cost with size molecular system. In response, there has been a surge interest in leveraging artificial intelligence (AI) machine learning (ML) techniques silico experiments. Integrating AI ML into increases scalability speed exploration space. remain, particularly regarding reproducibility transferability models. This review highlights evolution from, complementing, or replacing energy property predictions. Starting from models trained entirely on numerical data, journey set forth toward ideal model incorporating physical laws quantum mechanics. paper also reviews existing their intertwining, outlines roadmap future research, identifies areas improvement innovation. Ultimately, goal develop architectures capable accurate transferable solutions equation, thereby revolutionizing experiments within materials science.

Language: Английский

---