A Bayesian approach to NMR crystal structure determination

Polymorph Identification for Flexible Molecules: Linear Regression Analysis of Experimental and Calculated Solution- and Solid-State NMR Data

The Δδ regression approach of Blade et al. [ J. Phys. Chem. A 2020, 124(43), 8959-8977] for accurately discriminating between solid forms using a combination of experimental solution- and solid-state NMR data with density functional theory (DFT) calculation is here extended to molecules with multiple conformational degrees of freedom, using furosemide polymorphs as an exemplar. As before, the differences in measured 1H and 13C chemical shifts between solution-state NMR and solid-state magic-angle spinning (MAS) NMR (Δδexperimental) are compared to those determined by gauge-including projector augmented wave (GIPAW) calculations (Δδcalculated) by regression analysis and a t-test, allowing the correct furosemide polymorph to be precisely identified. Monte Carlo random sampling is used to calculate solution-state NMR chemical shifts, reducing computation times by avoiding the need to systematically sample the multidimensional conformational landscape that furosemide occupies in solution. The solvent conditions should be chosen to match the molecule's charge state between the solution and solid states. The Δδ regression approach indicates whether or not correlations between Δδexperimental and Δδcalculated are statistically significant; the approach is differently sensitive to the popular root mean squared error (RMSE) method, being shown to exhibit a much greater dynamic range. An alternative method for estimating solution-state NMR chemical shifts by approximating the measured solution-state dynamic 3D behavior with an ensemble of 54 furosemide crystal structures (polymorphs and cocrystals) from the Cambridge Structural Database (CSD) was also successful in this case, suggesting new avenues for this method that may overcome its current dependency on the prior determination of solution dynamic 3D structures.

An unusual ionic cocrystal of ponatinib hydrochloride: characterization by single-crystal X-ray diffraction and ultra-high field NMR spectroscopy

This study describes the discovery of a unique ionic cocrystal of the active pharmaceutical ingredient (API) ponatinib hydrochloride (pon·HCl), and characterization using single-crystal X-ray diffraction (SCXRD) and solid-state NMR (SSNMR) spectroscopy. Pon·HCl is a multicomponent crystal that features an unusual stoichiometry, with an asymmetric unit containing both monocations and dications of the ponatinib molecule, three water molecules, and three chloride ions. Structural features include (i) a charged imidazopyridazine moiety that forms a hydrogen bond between the ponatinib monocations and dications and (ii) a chloride ion that does not feature hydrogen bonds involving any organic moiety, instead being situated in a "square" arrangement with three water molecules. Multinuclear SSNMR, featuring high and ultra-high fields up to 35.2 T, provides the groundwork for structural interpretation of complex multicomponent crystals in the absence of diffraction data. A 13C CP/MAS spectrum confirms the presence of two crystallographically distinct ponatinib molecules, whereas 1D 1H and 2D 1H-1H DQ-SQ spectra identify and assign the unusually deshielded imidazopyridazine proton. 1D 35Cl spectra obtained at multiple fields confirm the presence of three distinct chloride ions, with density functional theory calculations providing key relationships between the SSNMR spectra and H⋯Cl- hydrogen bonding arrangements. A 2D 35Cl → 1H D-RINEPT spectrum confirms the spatial proximities between the chloride ions, water molecules, and amine moieties. This all suggests future application of multinuclear SSNMR at high and ultra-high fields to the study of complex API solid forms for which SCXRD data are unavailable, with potential application to heterogeneous mixtures or amorphous solid dispersions.

Advances in the Application of Artificial Intelligence-Based Spectral Data Interpretation: A Perspective

The interpretation of spectral data, including mass, nuclear magnetic resonance, infrared, and ultraviolet-visible spectra, is critical for obtaining molecular structural information. The development of advanced sensing technology has multiplied the amount of available spectral data. Chemical experts must use basic principles corresponding to the spectral information generated by molecular fragments and functional groups. This is a time-consuming process that requires a solid professional knowledge base. In recent years, the rapid development of computer science and its applications in cheminformatics and the emergence of computer-aided expert systems have greatly reduced the difficulty in analyzing large quantities of data. For expert systems, however, the problem-solving strategy must be known in advance or extracted by human experts and translated into algorithms. Gratifyingly, the development of artificial intelligence (AI) methods has shown great promise for solving such problems. Traditional algorithms, including the latest neural network algorithms, have shown great potential for both extracting useful information and processing massive quantities of data. This Perspective highlights recent innovations covering all of the emerging AI-based spectral interpretation techniques. In addition, the main limitations and current obstacles are presented, and the corresponding directions for further research are proposed. Moreover, this Perspective gives the authors' personal outlook on the development and future applications of spectral interpretation.

Federated Learning in Computational Toxicology: An Industrial Perspective on the Effiris Hackathon

In silico approaches have acquired a towering role in pharmaceutical research and development, allowing laboratories all around the world to design, create, and optimize novel molecular entities with unprecedented efficiency. From a toxicological perspective, computational methods have guided the choices of medicinal chemists toward compounds displaying improved safety profiles. Even if the recent advances in the field are significant, many challenges remain active in the on-target and off-target prediction fields. Machine learning methods have shown their ability to identify molecules with safety concerns. However, they strongly depend on the abundance and diversity of data used for their training. Sharing such information among pharmaceutical companies remains extremely limited due to confidentiality reasons, but in this scenario, a recent concept named "federated learning" can help overcome such concerns. Within this framework, it is possible for companies to contribute to the training of common machine learning algorithms, using, but not sharing, their proprietary data. Very recently, Lhasa Limited organized a hackathon involving several industrial partners in order to assess the performance of their federated learning platform, called "Effiris". In this paper, we share our experience as Roche in participating in such an event, evaluating the performance of the federated algorithms and comparing them with those coming from our in-house-only machine learning models. Our aim is to highlight the advantages of federated learning and its intrinsic limitations and also suggest some points for potential improvements in the method.

Discovering the Solid-State Secrets of Lorlatinib by NMR Crystallography: To Hydrogen Bond or not to Hydrogen Bond

Lorlatinib is an active pharmaceutical ingredient (API) used in the treatment of lung cancer. Here, an NMR crystallography analysis is presented whereby the single-crystal X-ray diffraction structure (CSD: 2205098) determination is complemented by multinuclear (1H, 13C, 14/15N, 19F) magic-angle spinning (MAS) solid-state NMR and gauge-including projector augmented wave (GIPAW) calculation of NMR chemical shifts. Lorlatinib crystallises in the P21 space group, with two distinct molecules in the asymmetric unit cell, Z' = 2. Three of the four NH2 hydrogen atoms form intermolecular hydrogen bonds, N30-H…N15 between the two distinct molecules and N30-H…O2 between two equivalent molecules. This is reflected in one of the NH21H chemical shifts being significantly lower, 4.0 ppm compared to 7.0 ppm. Two-dimensional 1H-13C, 14N-1H and 1H (double-quantum, DQ)-1H (single-quantum, SQ) MAS NMR spectra are presented. The 1H resonances are assigned and specific HH proximities corresponding to the observed DQ peaks are identified. The resolution enhancement at a 1H Larmor frequency of 1 GHz as compared to 500 or 600 MHz is demonstrated.

Quantum Chemical Roots of Machine-Learning Molecular Similarity Descriptors

In this work, we explore the quantum chemical foundations of descriptors for molecular similarity. Such descriptors are key for traversing chemical compound space with machine learning. Our focus is on the Coulomb matrix and on the smooth overlap of atomic positions (SOAP). We adopt a basic framework that allows us to connect both descriptors to electronic structure theory. This framework enables us to then define two new descriptors that are more closely related to electronic structure theory, which we call Coulomb lists and smooth overlap of electron densities (SOED). By investigating their usefulness as molecular similarity descriptors, we gain new insights into how and why Coulomb matrix and SOAP work. Moreover, Coulomb lists avoid the somewhat mysterious diagonalization step of the Coulomb matrix and might provide a direct means to extract subsystem information that can be compared across Born-Oppenheimer surfaces of varying dimension. For the electron density, we derive the necessary formalism to create the SOED measure in close analogy to SOAP. Because this formalism is more involved than that of SOAP, we review the essential theory as well as introduce a set of approximations that eventually allow us to work with SOED in terms of the same implementation available for the evaluation of SOAP. We focus our analysis on elementary reaction steps, where transition state structures are more similar to either reactant or product structures than the latter two are with respect to one another. The prediction of electronic energies of transition state structures can, however, be more difficult than that of stable intermediates due to multi-configurational effects. The question arises to what extent molecular similarity descriptors rooted in electronic structure theory can resolve these intricate effects.

A Machine Learning Model of Chemical Shifts for Chemically and Structurally Diverse Molecular Solids

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.

Quantum machine learning for chemistry and physics

Machine learning (ML) has emerged as a formidable force for identifying hidden but pertinent patterns within a given data set with the objective of subsequent generation of automated predictive behavior. In recent years, it is safe to conclude that ML and its close cousin, deep learning (DL), have ushered in unprecedented developments in all areas of physical sciences, especially chemistry. Not only classical variants of ML, even those trainable on near-term quantum hardwares have been developed with promising outcomes. Such algorithms have revolutionized materials design and performance of photovoltaics, electronic structure calculations of ground and excited states of correlated matter, computation of force-fields and potential energy surfaces informing chemical reaction dynamics, reactivity inspired rational strategies of drug designing and even classification of phases of matter with accurate identification of emergent criticality. In this review we shall explicate a subset of such topics and delineate the contributions made by both classical and quantum computing enhanced machine learning algorithms over the past few years. We shall not only present a brief overview of the well-known techniques but also highlight their learning strategies using statistical physical insight. The objective of the review is not only to foster exposition of the aforesaid techniques but also to empower and promote cross-pollination among future research in all areas of chemistry which can benefit from ML and in turn can potentially accelerate the growth of such algorithms.

Solid-state multinuclear magnetic resonance and X-ray crystallographic investigation of the phosphorus...iodine halogen bond in a bis(dicyclohexylphenylphosphine)(1,6-diiodoperfluorohexane) cocrystal

Halogen bonding to phosphorus atoms remains uncommon, with relatively few examples reported in the literature. Here, the preparation and investigation of the cocrystal bis(dicyclohexylphenylphosphine)(1,6-diiodoperfluorohexane) by X-ray crystallography and solid-state multinuclear magnetic resonance spectroscopy is described. The crystal structure features two crystallographically unique C-I...P halogen bonds [d_I...P = 3.090 (5) Å, 3.264 (5) Å] and crystallographic disorder of one of the 1,6-diiodoperfluorohexane molecules. The first of these is the shortest and most linear I...P halogen bond reported to date. ¹³C, ¹⁹F, and ³¹P magic angle spinning solid-state NMR spectra are reported. A ³¹P chemical shift change of -7.0 p.p.m. in the cocrystal relative to pure dicyclohexylphenylphosphine, consistent with halogen bond formation, is noted. This work establishes iodoperfluoroalkanes as viable halogen bond donors when paired with phosphorus acceptors, and also shows that dicyclohexylphenylphosphine can act as a practical halogen bond acceptor.

De Novo Crystal Structure Determination from Machine Learned Chemical Shifts

Determination of the three-dimensional atomic-level structure of powdered solids is one of the key goals in current chemistry. Solid-state NMR chemical shifts can be used to solve this problem, but they are limited by the high computational cost associated with crystal structure prediction methods and density functional theory chemical shift calculations. Here, we successfully determine the crystal structures of ampicillin, piroxicam, cocaine, and two polymorphs of the drug molecule AZD8329 using on-the-fly generated machine-learned isotropic chemical shifts to directly guide a Monte Carlo-based structure determination process starting from a random gas-phase conformation.

Bayesian Inference Applied to NMR-Based Configurational Assignments by Floating Chirality Distance Geometry Calculations

Using NMR data, the assignment of the correct 3D configuration and conformation to unknown natural products is of pivotal importance in pharmaceutical and medicinal chemistry. In this report, we quantify the quality and probability of structural elucidations using Bayesian inference in combination with floating chirality distance geometry simulations. Here, we will discuss the configurational analysis of three complex natural products including isopinocampheol (1), plakilactone H (2), and iodocallophycoic acid A (3) using NMR restraints of various types and in different combinations (residual dipolar couplings (RDCs) and NOE-derived distances). Our results quantitatively demonstrate how reliably molecular geometries can be inferred from experimental NMR data, unequivocally unveiling remaining assignment ambiguities. The methodology presented here can dramatically reduce the risk of incorrect structural assignments based on the overinterpretation of incomplete data in chemistry.

Atomic-resolution chemical characterization of (2x)72-kDa tryptophan synthase via four- and five-dimensional 1H-detected solid-state NMR

NMR chemical shifts provide detailed information on the chemical properties of molecules, thereby complementing structural data from techniques like X-ray crystallography and electron microscopy. Detailed analysis of protein NMR data, however, often hinges on comprehensive, site-specific assignment of backbone resonances, which becomes a bottleneck for molecular weights beyond 40 to 45 kDa. Here, we show that assignments for the (2x)72-kDa protein tryptophan synthase (665 amino acids per asymmetric unit) can be achieved via higher-dimensional, proton-detected, solid-state NMR using a single, 1-mg, uniformly labeled, microcrystalline sample. This framework grants access to atom-specific characterization of chemical properties and relaxation for the backbone and side chains, including those residues important for the catalytic turnover. Combined with first-principles calculations, the chemical shifts in the β-subunit active site suggest a connection between active-site chemistry, the electrostatic environment, and catalytically important dynamics of the portal to the β-subunit from solution.

Imaging active site chemistry and protonation states: NMR crystallography of the tryptophan synthase α-aminoacrylate intermediate

NMR-assisted crystallography-the integrated application of solid-state NMR, X-ray crystallography, and first-principles computational chemistry-holds significant promise for mechanistic enzymology: by providing atomic-resolution characterization of stable intermediates in enzyme active sites, including hydrogen atom locations and tautomeric equilibria, NMR crystallography offers insight into both structure and chemical dynamics. Here, this integrated approach is used to characterize the tryptophan synthase α-aminoacrylate intermediate, a defining species for pyridoxal-5'-phosphate-dependent enzymes that catalyze β-elimination and replacement reactions. For this intermediate, NMR-assisted crystallography is able to identify the protonation states of the ionizable sites on the cofactor, substrate, and catalytic side chains as well as the location and orientation of crystallographic waters within the active site. Most notable is the water molecule immediately adjacent to the substrate β-carbon, which serves as a hydrogen bond donor to the ε-amino group of the acid-base catalytic residue βLys87. From this analysis, a detailed three-dimensional picture of structure and reactivity emerges, highlighting the fate of the L-serine hydroxyl leaving group and the reaction pathway back to the preceding transition state. Reaction of the α-aminoacrylate intermediate with benzimidazole, an isostere of the natural substrate indole, shows benzimidazole bound in the active site and poised for, but unable to initiate, the subsequent bond formation step. When modeled into the benzimidazole position, indole is positioned with C3 in contact with the α-aminoacrylate Cβ and aligned for nucleophilic attack. Here, the chemically detailed, three-dimensional structure from NMR-assisted crystallography is key to understanding why benzimidazole does not react, while indole does.

NMR-Based Configurational Assignments of Natural Products: Gibbs Sampling and Bayesian Inference Using Floating Chirality Distance Geometry Calculations

Floating chirality restrained distance geometry (fc-rDG) calculations are used to directly evolve structures from NMR data such as NOE-derived intramolecular distances or anisotropic residual dipolar couplings (RDCs). In contrast to evaluating pre-calculated structures against NMR restraints, multiple configurations (diastereomers) and conformations are generated automatically within the experimental limits. In this report, we show that the "unphysical" rDG pseudo energies defined from NMR violations bear statistical significance, which allows assigning probabilities to configurational assignments made that are fully compatible with the method of Bayesian inference. These "diastereomeric differentiabilities" then even become almost independent of the actual values of the force constants used to model the restraints originating from NOE or RDC data.

A toolbox for improving the workflow of NMR crystallography

NMR crystallography is a powerful tool with applications in structural characterization and crystal structure verification, to name two. However, applying this tool presents several challenges, especially for industrial users, in terms of consistency, workflow, time consumption, and the requirement for a high level of understanding of experimental solid-state NMR and GIPAW-DFT calculations. Here, we have developed a series of fully parameterized scripts for use in Materials Studio and TopSpin, based on the .magres file format, with a focus on organic molecules (e.g. pharmaceuticals), improving efficiency, robustness, and workflow. We separate these tools into three major categories: performing the DFT calculations, extracting & visualizing the results, and crystallographic modelling. These scripts will rapidly submit fully parameterized CASTEP jobs, extract data from the calculations, assist in visualizing the results, and expedite the process of structural modelling. Accompanied with these tools is a description on their functionality, documentation on how to get started and use the scripts, and links to video tutorials for guiding new users. Through the use of these tools, we hope to facilitate NMR crystallography and to harmonize the process across users.

Bayesian probabilistic assignment of chemical shifts in organic solids

A prerequisite for NMR studies of organic materials is assigning each experimental chemical shift to a set of geometrically equivalent nuclei. Obtaining the assignment experimentally can be challenging and typically requires time-consuming multidimensional correlation experiments. An alternative solution for determining the assignment involves statistical analysis of experimental chemical shift databases, but no such database exists for molecular solids. Here, by combining the Cambridge Structural Database with a machine learning model of chemical shifts, we construct a statistical basis for probabilistic chemical shift assignment of organic crystals by calculating shifts for more than 200,000 compounds, enabling the probabilistic assignment of organic crystals directly from their two-dimensional chemical structure. The approach is demonstrated with the ¹³C and ¹H assignment of 11 molecular solids with experimental shifts and benchmarked on 100 crystals using predicted shifts. The correct assignment was found among the two most probable assignments in more than 80% of cases.

Gaussian Process Regression for Materials and Molecules

We provide an introduction to Gaussian process regression (GPR) machine-learning methods in computational materials science and chemistry. The focus of the present review is on the regression of atomistic properties: in particular, on the construction of interatomic potentials, or force fields, in the Gaussian Approximation Potential (GAP) framework; beyond this, we also discuss the fitting of arbitrary scalar, vectorial, and tensorial quantities. Methodological aspects of reference data generation, representation, and regression, as well as the question of how a data-driven model may be validated, are reviewed and critically discussed. A survey of applications to a variety of research questions in chemistry and materials science illustrates the rapid growth in the field. A vision is outlined for the development of the methodology in the years to come.

Ab Initio Machine Learning in Chemical Compound Space

Chemical compound space (CCS), the set of all theoretically conceivable combinations of chemical elements and (meta-)stable geometries that make up matter, is colossal. The first-principles based virtual sampling of this space, for example, in search of novel molecules or materials which exhibit desirable properties, is therefore prohibitive for all but the smallest subsets and simplest properties. We review studies aimed at tackling this challenge using modern machine learning techniques based on (i) synthetic data, typically generated using quantum mechanics based methods, and (ii) model architectures inspired by quantum mechanics. Such Quantum mechanics based Machine Learning (QML) approaches combine the numerical efficiency of statistical surrogate models with an ab initio view on matter. They rigorously reflect the underlying physics in order to reach universality and transferability across CCS. While state-of-the-art approximations to quantum problems impose severe computational bottlenecks, recent QML based developments indicate the possibility of substantial acceleration without sacrificing the predictive power of quantum mechanics.

Importance of Nuclear Quantum Effects for NMR Crystallography

The resolving power of solid-state nuclear magnetic resonance (NMR) crystallography depends heavily on the accuracy of computational predictions of NMR chemical shieldings of candidate structures, which are usually taken to be local minima in the potential energy. To test the limits of this approximation, we systematically study the importance of finite-temperature and quantum nuclear fluctuations for ¹H, ¹³C, and ¹⁵N shieldings in polymorphs of three paradigmatic molecular crystals: benzene, glycine, and succinic acid. The effect of quantum fluctuations is comparable to the typical errors of shielding predictions for static nuclei with respect to experiments, and their inclusion improves the agreement with measurements, translating to more reliable assignment of the NMR spectra to the correct candidate structure. The use of integrated machine-learning models, trained on first-principles energies and shieldings, renders rigorous sampling of nuclear fluctuations affordable, setting a new standard for the calculations underlying NMR structure determinations.

Physics-Inspired Structural Representations for Molecules and Materials

The first step in the construction of a regression model or a data-driven analysis, aiming to predict or elucidate the relationship between the atomic-scale structure of matter and its properties, involves transforming the Cartesian coordinates of the atoms into a suitable representation. The development of atomic-scale representations has played, and continues to play, a central role in the success of machine-learning methods for chemistry and materials science. This review summarizes the current understanding of the nature and characteristics of the most commonly used structural and chemical descriptions of atomistic structures, highlighting the deep underlying connections between different frameworks and the ideas that lead to computationally efficient and universally applicable models. It emphasizes the link between properties, structures, their physical chemistry, and their mathematical description, provides examples of recent applications to a diverse set of chemical and materials science problems, and outlines the open questions and the most promising research directions in the field.

Structure determination of an amorphous drug through large-scale NMR predictions

Knowledge of the structure of amorphous solids can direct, for example, the optimization of pharmaceutical formulations, but atomic-level structure determination in amorphous molecular solids has so far not been possible. Solid-state nuclear magnetic resonance (NMR) is among the most popular methods to characterize amorphous materials, and molecular dynamics (MD) simulations can help describe the structure of disordered materials. However, directly relating MD to NMR experiments in molecular solids has been out of reach until now because of the large size of these simulations. Here, using a machine learning model of chemical shifts, we determine the atomic-level structure of the hydrated amorphous drug AZD5718 by combining dynamic nuclear polarization-enhanced solid-state NMR experiments with predicted chemical shifts for MD simulations of large systems. From these amorphous structures we then identify H-bonding motifs and relate them to local intermolecular complex formation energies.

Revving up 13C NMR shielding predictions across chemical space: benchmarks for atoms-in-molecules kernel machine learning with new data for 134 kilo molecules

The requirement for accelerated and quantitatively accurate screening of nuclear magnetic resonance spectra across the small molecules chemical compound space is two-fold: (1) a robust ‘local’ machine learning (ML) strategy capturing the effect of the neighborhood on an atom’s ‘near-sighted’ property—chemical shielding; (2) an accurate reference dataset generated with a state-of-the-art first-principles method for training. Herein we report the QM9-NMR dataset comprising isotropic shielding of over 0.8 million C atoms in 134k molecules of the QM9 dataset in gas and five common solvent phases. Using these data for training, we present benchmark results for the prediction transferability of kernel-ridge regression models with popular local descriptors. Our best model, trained on 100k samples, accurately predicts isotropic shielding of 50k ‘hold-out’ atoms with a mean error of less than 1.9 ppm. For the rapid prediction of new query molecules, the models were trained on geometries from an inexpensive theory. Furthermore, by using a Δ-ML strategy, we quench the error below 1.4 ppm. Finally, we test the transferability on non-trivial benchmark sets that include benchmark molecules comprising 10–17 heavy atoms and drugs.

Solid-state NMR spectroscopy

Solid-state nuclear magnetic resonance (NMR) spectroscopy is an atomic-level method used to determine the chemical structure, three-dimensional structure, and dynamics of solids and semi-solids. This Primer summarizes the basic principles of NMR as applied to the wide range of solid systems. The fundamental nuclear spin interactions and the effects of magnetic fields and radiofrequency pulses on nuclear spins are the same as in liquid-state NMR. However, because of the anisotropy of the interactions in the solid state, the majority of high-resolution solid-state NMR spectra is measured under magic-angle spinning (MAS), which has profound effects on the types of radiofrequency pulse sequences required to extract structural and dynamical information. We describe the most common MAS NMR experiments and data analysis approaches for investigating biological macromolecules, organic materials, and inorganic solids. Continuing development of sensitivity-enhancement approaches, including 1H-detected fast MAS experiments, dynamic nuclear polarization, and experiments tailored to ultrahigh magnetic fields, is described. We highlight recent applications of solid-state NMR to biological and materials chemistry. The Primer ends with a discussion of current limitations of NMR to study solids, and points to future avenues of development to further enhance the capabilities of this sophisticated spectroscopy for new applications.

Towards Accurate Predictions of Proton NMR Spectroscopic Parameters in Molecular Solids

The factors contributing to the accuracy of quantum-chemical calculations for the prediction of proton NMR chemical shifts in molecular solids are systematically investigated. Proton chemical shifts of six solid amino acids with hydrogen atoms in various bonding environments (CH, CH₂ , CH₃ , OH, SH and NH₃ ) were determined experimentally using ultra-fast magic-angle spinning and proton-detected 2D NMR experiments. The standard DFT method commonly used for the calculations of NMR parameters of solids is shown to provide chemical shifts that deviate from experiment by up to 1.5 ppm. The effects of the computational level (hybrid DFT functional, coupled-cluster calculation, inclusion of relativistic spin-orbit coupling) are thoroughly discussed. The effect of molecular dynamics and nuclear quantum effects are investigated using path-integral molecular dynamics (PIMD) simulations. It is demonstrated that the accuracy of the calculated proton chemical shifts is significantly better when these effects are included in the calculations.

Investigating discrepancies between experimental solid-state NMR and GIPAW calculation: NC-N 13C and OH⋯O 1H chemical shifts in pyridinium fumarates and their cocrystals

An NMR crystallography analysis is presented for four solid-state structures of pyridine fumarates and their cocrystals, using crystal structures deposited in the Cambridge Crystallographic Data Centre, CCDC. Experimental one-dimensional one-pulse ¹H and ¹³C cross-polarisation (CP) magic-angle spinning (MAS) nuclear magnetic resonance (NMR) and two-dimensional ¹⁴N-¹H heteronuclear multiple-quantum coherence MAS NMR spectra are compared with gauge-including projector augmented wave (GIPAW) calculations of the ¹H and ¹³C chemical shifts and the ¹⁴N shifts that additionally depend on the quadrupolar interaction. Considering the high ppm (>10 ​ppm) ¹H resonances, while there is good agreement (within 0.4 ​ppm) between experiment and GIPAW calculation for the hydrogen-bonded NH moieties, the hydrogen-bonded fumaric acid OH resonances are 1.2-1.9 ​ppm higher in GIPAW calculation as compared to experiment. For the cocrystals of a salt and a salt formed by 2-amino-5-methylpyridinium and 2-amino-6-methylpyridinium ions, a large discrepancy of 4.2 and 5.9 ​ppm between experiment and GIPAW calculation is observed for the quaternary ring carbon ¹³C resonance that is directly bonded to two nitrogens (in the ring and in the amino group). By comparison, there is excellent agreement (within 0.2 ​ppm) for the quaternary ring carbon ¹³C resonance directly bonded to the ring nitrogen for the salt and cocrystal of a salt formed by 2,6-lutidinium and 2,5-lutidinium, respectively.

PyShifts: A PyMOL Plugin for Chemical Shift-Based Analysis of Biomolecular Ensembles

Here, we present PyShifts-a PyMOL plugin for chemical shift-based analysis of biomolecular ensembles. With PyShifts, users can compare and visualize differences between experimentally measured and computationally predicted chemical shifts. When analyzing multiple conformations of a biomolecule with PyShifts, users can also sort a set of conformations based on chemical shift differences and identify the conformers that exhibit the best agreement between measured and predicted chemical shifts. Although we have integrated PyShifts with the chemical shift predictors LARMOR^D and LARMOR^Cα, PyShifts can read in chemical shifts from any source, and so, users can employ PyShifts to analyze biomolecular structures using chemical shifts computed by any chemical shift predictor. We envision, therefore, that PyShifts (https://github.com/atfrank/PyShifts) will find utility as a general-purpose tool for exploring chemical shift-structure relationships in biomolecular ensembles.

Quantum Chemistry in the Age of Machine Learning

As the quantum chemistry (QC) community embraces machine learning (ML), the number of new methods and applications based on the combination of QC and ML is surging. In this Perspective, a view of the current state of affairs in this new and exciting research field is offered, challenges of using machine learning in quantum chemistry applications are described, and potential future developments are outlined. Specifically, examples of how machine learning is used to improve the accuracy and accelerate quantum chemical research are shown. Generalization and classification of existing techniques are provided to ease the navigation in the sea of literature and to guide researchers entering the field. The emphasis of this Perspective is on supervised machine learning.