Machine Learning of Reactive Potentials.

In the past two decades, machine learning potentials (MLPs) have driven significant developments in chemical, biological, and material sciences. The construction and training of MLPs enable fast and accurate simulations and analysis of thermodynamic and kinetic properties. This review focuses on the application of MLPs to reaction systems with consideration of bond breaking and formation. We review the development of MLP models, primarily with neural network and kernel-based algorithms, and recent applications of reactive MLPs (RMLPs) to systems at different scales. We show how RMLPs are constructed, how they speed up the calculation of reactive dynamics, and how they facilitate the study of reaction trajectories, reaction rates, free energy calculations, and many other calculations. Different data sampling strategies applied in building RMLPs are also discussed with a focus on how to collect structures for rare events and how to further improve their performance with active learning.

Assessment of Embedding Schemes in a Hybrid Machine Learning/Classical Potentials (ML/MM) Approach.

Machine learning (ML) methods have reached high accuracy levels for the prediction of in vacuo molecular properties. However, the simulation of large systems solely through ML methods (such as those based on neural network potentials) is still a challenge. In this context, one of the most promising frameworks for integrating ML schemes in the simulation of complex molecular systems are the so-called ML/MM methods. These multiscale approaches combine ML methods with classical force fields (MM), in the same spirit as the successful hybrid quantum mechanics-molecular mechanics methods (QM/MM). The key issue for such ML/MM methods is an adequate description of the coupling between the region of the system described by ML and the region described at the MM level. In the context of QM/MM schemes, the main ingredient of the interaction is electrostatic, and the state of the art is the so-called electrostatic-embedding. In this study, we analyze the quality of simpler mechanical embedding-based approaches, specifically focusing on their application within a ML/MM framework utilizing atomic partial charges derived in vacuo. Taking as reference electrostatic embedding calculations performed at a QM(DFT)/MM level, we explore different atomic charges schemes, as well as a polarization correction computed using atomic polarizabilites. Our benchmark data set comprises a set of about 80k small organic structures from the ANI-1x and ANI-2x databases, solvated in water. The results suggest that the minimal basis iterative stockholder (MBIS) atomic charges yield the best agreement with the reference coupling energy. Remarkable enhancements are achieved by including a simple polarization correction.

Axially Chiral Cannabinoids: Design, Synthesis, and Cannabinoid Receptor Affinity.

The resorcinol-terpene phytocannabinoid template is a privileged scaffold for the development of diverse therapeutics targeting the endocannabinoid system. Axially chiral cannabinols (axCBNs) are unnatural cannabinols (CBNs) that bear an additional C10 substituent, which twists the cannabinol biaryl framework out of planarity creating an axis of chirality. This unique structural modification is hypothesized to enhance both the physical and biological properties of cannabinoid ligands, thus ushering in the next generation of endocannabinoid system chemical probes and cannabinoid-inspired leads for drug development. In this full report, we describe the philosophy guiding the design of axCBNs as well as several synthetic strategies for their construction. We also introduce a second class of axially chiral cannabinoids inspired by cannabidiol (CBD), termed axially chiral cannabidiols (axCBDs). Finally, we provide an analysis of axially chiral cannabinoid (axCannabinoid) atropisomerism, which spans two classes (class 1 and 3 atropisomers), and provide first evidence that axCannabinoids retain─and in some cases, strengthen─affinity and functional activity at cannabinoid receptors. Together, these findings present a promising new direction for the design of novel cannabinoid ligands for drug discovery and exploration of the complex endocannabinoid system.

Novel Lennard-Jones Parameters for Cysteine and Selenocysteine in the AMBER Force Field.

Cysteine is a common amino acid with a thiol group that plays a pivotal role in a variety of scenarios in redox biochemistry. In contrast, selenocysteine, the 21st amino acid, is only present in 25 human proteins. Classical force-field parameters for cysteine and selenocysteine are still scarce. In this context, we present a methodology to obtain Lennard-Jones parameters for cysteine and selenocysteine in different physiologically relevant oxidation and protonation states. The new force field parameters obtained in this work are available at https://github.com/MALBECC/AMBER-parameters-database. The parameters were adjusted to reproduce water radial distribution functions obtained by density functional theory ab initio molecular dynamics. We validated the results by evaluating the impact of the choice of parameters on the structure and dynamics in classical molecular dynamics simulations of representative proteins containing catalytic cysteine/selenocysteine residues. There are significant changes in protein structure and dynamics depending on the parameters choice, specifically affecting the residues close to the catalytic sites.

AmberTools.

AmberTools is a free and open-source collection of programs used to set up, run, and analyze molecular simulations. The newer features contained within AmberTools23 are briefly described in this Application note.

Auto3D: Automatic Generation of the Low-Energy 3D Structures with ANI Neural Network Potentials.

Computational programs accelerate the chemical discovery processes but often need proper three-dimensional molecular information as part of the input. Getting optimal molecular structures is challenging because it requires enumerating and optimizing a huge space of stereoisomers and conformers. We developed the Python-based Auto3D package for generating the low-energy 3D structures using SMILES as the input. Auto3D is based on state-of-the-art algorithms and can automatize the isomer enumeration and duplicate filtering process, 3D building process, geometry optimization, and ranking process. Tested on 50 molecules with multiple unspecified stereocenters, Auto3D is guaranteed to find the stereoconfiguration that yields the lowest-energy conformer. With Auto3D, we provide an extension of the ANI model. The new model, dubbed ANI-2xt, is trained on a tautomer-rich data set. ANI-2xt is benchmarked with DFT methods on geometry optimization and electronic and Gibbs free energy calculations. Compared with ANI-2x, ANI-2xt provides a 42% error reduction for tautomeric reaction energy calculations when using the gold-standard coupled-cluster calculation as the reference. ANI-2xt can accurately predict the energies and is several orders of magnitude faster than DFT methods.

Non-adiabatic Excited-State Molecular Dynamics: Theory and Applications for Modeling Photophysics in Extended Molecular Materials.

Optically active molecular materials, such as organic conjugated polymers and biological systems, are characterized by strong coupling between electronic and vibrational degrees of freedom. Typically, simulations must go beyond the Born-Oppenheimer approximation to account for non-adiabatic coupling between excited states. Indeed, non-adiabatic dynamics is commonly associated with exciton dynamics and photophysics involving charge and energy transfer, as well as exciton dissociation and charge recombination. Understanding the photoinduced dynamics in such materials is vital to providing an accurate description of exciton formation, evolution, and decay. This interdisciplinary field has matured significantly over the past decades. Formulation of new theoretical frameworks, development of more efficient and accurate computational algorithms, and evolution of high-performance computer hardware has extended these simulations to very large molecular systems with hundreds of atoms, including numerous studies of organic semiconductors and biomolecules. In this Review, we will describe recent theoretical advances including treatment of electronic decoherence in surface-hopping methods, the role of solvent effects, trivial unavoided crossings, analysis of data based on transition densities, and efficient computational implementations of these numerical methods. We also emphasize newly developed semiclassical approaches, based on the Gaussian approximation, which retain phase and width information to account for significant decoherence and interference effects while maintaining the high efficiency of surface-hopping approaches. The above developments have been employed to successfully describe photophysics in a variety of molecular materials.

QSAR without borders.

Prediction of chemical bioactivity and physical properties has been one of the most important applications of statistical and more recently, machine learning and artificial intelligence methods in chemical sciences. This field of research, broadly known as quantitative structure-activity relationships (QSAR) modeling, has developed many important algorithms and has found a broad range of applications in physical organic and medicinal chemistry in the past 55+ years. This Perspective summarizes recent technological advances in QSAR modeling but it also highlights the applicability of algorithms, modeling methods, and validation practices developed in QSAR to a wide range of research areas outside of traditional QSAR boundaries including synthesis planning, nanotechnology, materials science, biomaterials, and clinical informatics. As modern research methods generate rapidly increasing amounts of data, the knowledge of robust data-driven modelling methods professed within the QSAR field can become essential for scientists working both within and outside of chemical research. We hope that this contribution highlighting the generalizable components of QSAR modeling will serve to address this challenge.

Correction: QSAR without borders.

Correction for 'QSAR without borders' by Eugene N. Muratov et al., Chem. Soc. Rev., 2020, DOI: 10.1039/d0cs00098a.

Exploring Coupled Redox and pH Processes with a Force-Field-Based Approach: Applications to Five Different Systems.

Coupled redox and pH-driven processes are at the core of many important biological mechanisms. As the distribution of protonation and redox states in a system is associated with the pH and redox potential of the solution, having efficient computational tools that can simulate under these conditions becomes very important. Such tools have the potential to provide information that complement and drive experiments. In previous publications we have presented the implementation of the constant pH and redox potential molecular dynamics (C(pH,E)MD) method in AMBER and we have shown how multidimensional replica exchange can be used to significantly enhance the convergence efficiency of our simulations. In the current work, after an improvement in our C(pH,E)MD approach that allows a given residue to be simultaneously pH- and redox-active, we have employed our methodologies to study five different systems of interest in the literature. We present results for capped tyrosine dipeptide, two maquette systems containing one pH- and redox-active tyrosine (α3Y and peptide A), and two proteins that contain multiple heme groups (diheme cytochrome c from Rhodobacter sphaeroides and Desulfovibrio vulgaris Hildenborough cytochrome c3). We show that our results can provide new insights into previous theoretical and experimental findings by using a fully force-field-based and GPU-accelerated approach, which allows the simulations to be executed with high computational performance.

TorchANI: A Free and Open Source PyTorch-Based Deep Learning Implementation of the ANI Neural Network Potentials.

This paper presents TorchANI, a PyTorch-based program for training/inference of ANI (ANAKIN-ME) deep learning models to obtain potential energy surfaces and other physical properties of molecular systems. ANI is an accurate neural network potential originally implemented using C++/CUDA in a program called NeuroChem. Compared with NeuroChem, TorchANI has a design emphasis on being lightweight, user friendly, cross platform, and easy to read and modify for fast prototyping, while allowing acceptable sacrifice on running performance. Because the computation of atomic environmental vectors and atomic neural networks are all implemented using PyTorch operators, TorchANI is able to use PyTorch's autograd engine to automatically compute analytical forces and Hessian matrices, as well as do force training without requiring any additional codes. TorchANI is open-source and freely available on GitHub: https://github.com/aiqm/torchani.

The generalized Boltzmann distribution is the only distribution in which the Gibbs-Shannon entropy equals the thermodynamic entropy.

We show that the generalized Boltzmann distribution is the only distribution for which the Gibbs-Shannon entropy equals the thermodynamic entropy. This result means that the thermodynamic entropy and the Gibbs-Shannon entropy are not generally equal, but rather the equality holds only in the special case where a system is in equilibrium with a reservoir.

A Coupled Ionization-Conformational Equilibrium Is Required To Understand the Properties of Ionizable Residues in the Hydrophobic Interior of Staphylococcal Nuclease.

Ionizable residues in the interior of proteins play essential roles, especially in biological energy transduction, but are relatively rare and seem incompatible with the complex and polar environment. We perform a comprehensive study of the internal ionizable residues on 21 variants of staphylococcal nuclease with internal Lys, Glu, or Asp residues. Using pH replica exchange molecular dynamics simulations, we find that, in most cases, the pKa values of these internal ionizable residues are shifted significantly from their values in solution. Our calculated results are in excellent agreement with the experimental observations of the Garcia-Moreno group. We show that the interpretation of the experimental pKa values requires the study of not only protonation changes but also conformational changes. The coupling between the protonation and conformational equilibria suggests a mechanism for efficient pH-sensing and regulation in proteins. This study provides new physical insights into how internal ionizable residues behave in the hydrophobic interior of proteins.

GPU-Accelerated Molecular Dynamics and Free Energy Methods in Amber18: Performance Enhancements and New Features.

We report progress in graphics processing unit (GPU)-accelerated molecular dynamics and free energy methods in Amber18. Of particular interest is the development of alchemical free energy algorithms, including free energy perturbation and thermodynamic integration methods with support for nonlinear soft-core potential and parameter interpolation transformation pathways. These methods can be used in conjunction with enhanced sampling techniques such as replica exchange, constant-pH molecular dynamics, and new 12-6-4 potentials for metal ions. Additional performance enhancements have been made that enable appreciable speed-up on GPUs relative to the previous software release.

Probing the Structures of Solvent-Complexed Ions Formed in Electrospray Ionization Using Cryogenic Infrared Photodissociation Spectroscopy.

The gas-phase infrared photodissociation (IRPD) spectra of solvent-tagged small biomolecules are studied in a cryogenic ion trap at 17 K. In this study para-aminobenzoic acid (PABA) and tyramine molecules are noncovalently tagged with water or acetonitrile in the electrospray ionization (ESI) source. The complexes are then cooled in the cryogenic trap prior to spectroscopic measurements. These molecules provide two putative sites for solvent attachment: the protonated amine (NH3+) and the OH groups. Comparisons of the experimental IR spectra to theoretical spectra obtained with density functional theory show that the NH3+ site is mainly favored. Evidence for the formation of both NH3-bound and OH-bound conformers is found only in tyramine, despite having similar solution- and gas-phase energetics to that of PABA. Since the structures cannot interconvert in the gas phase, this suggests an isomerization during the electrospray process.

Cation-Dependent Conformations in 25-Hydroxyvitamin D3-Cation Adducts Measured by Ion Mobility-Mass Spectrometry and Theoretical Modeling.

Ion mobility-mass spectrometry is a useful tool in separation of biological isomers, including clinically relevant analytes such as 25-hydroxyvitamin D3 (25OHD3) and its epimer, 3-epi-25-hydroxyvitamin D3 (epi25OHD3). Previous research indicates that these epimers adopt different gas-phase sodiated monomer structures, either the "open" or "closed" conformer, which allow 25OHD3 to be readily resolved in mixtures. In the current work, alternative metal cation adducts are investigated for their relative effects on the ratio of "open" and "closed conformers. Alkali and alkaline earth metal adducts caused changes in the 25OHD3 conformer ratio, where the proportion of the "open" conformer generally increases with the size of the metal cation in a given group. As such, the ratio of the "open" conformer, which is unique to 25OHD3 and absent for its epimer, can be increased from approximately 1:1 for the sodiated monomer to greater than 8:1 for the barium adduct. Molecular modeling and energy calculations agree with the experimental results, indicating that the Gibbs free energy of conversion from the "closed" to the "open" conformation decreased with increasing cation size, correlating with the variation in ratio between the conformers. This work demonstrates the effect of cation adducts on gas-phase conformations of small, flexible molecules and offers an additional strategy for resolution of clinically relevant epimers.

Redox potential replica exchange molecular dynamics at constant pH in AMBER: Implementation and validation.

Redox processes are important in chemistry, with applications in biomedicine, chemical analysis, among others. As many redox experiments are also performed at a fixed value of pH, having an efficient computational method to support experimental measures at both constant redox potential and pH is very important. Such computational techniques have the potential to validate experimental observations performed under these conditions and to provide additional information unachievable experimentally such as an atomic level description of macroscopic measures. We present the implementation of discrete redox and protonation states methods for constant redox potential Molecular Dynamics (CEMD), for coupled constant pH and constant redox potential MD (C(pH,E)MD), and for Replica Exchange MD along the redox potential dimension (E-REMD) on the AMBER software package. Validation results are presented for a small system that contains a single heme group: N-acetylmicroperoxidase-8 (NAcMP8) axially connected to a histidine peptide. The methods implemented allow one to make standard redox potential (Eo) predictions with the same easiness and accuracy as pKa predictions using the constant pH molecular dynamics and pH-REMD methods currently available on AMBER. In our simulations, we can correctly describe, in agreement also with theoretical predictions, the following behaviors: when a redox-active group is reduced, the pKa of a near pH-active group increases because it becomes easier for a proton to be attached; equivalently, when a pH-active group is protonated, the standard redox potential (Eo) of an adjacent redox-active group rises. Furthermore, our results also show that E-REMD is able to achieve faster statistical convergence than CEMD or C(pH,E)MD. Moreover, computational benchmarks using our methodologies show high-performance of GPU (Graphics Processing Unit) accelerated calculations in comparison to conventional CPU (Central Processing Unit) calculations.

Less is more: Sampling chemical space with active learning.

The development of accurate and transferable machine learning (ML) potentials for predicting molecular energetics is a challenging task. The process of data generation to train such ML potentials is a task neither well understood nor researched in detail. In this work, we present a fully automated approach for the generation of datasets with the intent of training universal ML potentials. It is based on the concept of active learning (AL) via Query by Committee (QBC), which uses the disagreement between an ensemble of ML potentials to infer the reliability of the ensemble's prediction. QBC allows the presented AL algorithm to automatically sample regions of chemical space where the ML potential fails to accurately predict the potential energy. AL improves the overall fitness of ANAKIN-ME (ANI) deep learning potentials in rigorous test cases by mitigating human biases in deciding what new training data to use. AL also reduces the training set size to a fraction of the data required when using naive random sampling techniques. To provide validation of our AL approach, we develop the COmprehensive Machine-learning Potential (COMP6) benchmark (publicly available on GitHub) which contains a diverse set of organic molecules. Active learning-based ANI potentials outperform the original random sampled ANI-1 potential with only 10% of the data, while the final active learning-based model vastly outperforms ANI-1 on the COMP6 benchmark after training to only 25% of the data. Finally, we show that our proposed AL technique develops a universal ANI potential (ANI-1x) that provides accurate energy and force predictions on the entire COMP6 benchmark. This universal ML potential achieves a level of accuracy on par with the best ML potentials for single molecules or materials, while remaining applicable to the general class of organic molecules composed of the elements CHNO.

Structure-Activity Relationships of Benzenesulfonamide-Based Inhibitors towards Carbonic Anhydrase Isoform Specificity.

Carbonic anhydrases (CAs) are implicated in a wide range of diseases, including the upregulation of isoforms CA IX and XII in many aggressive cancers. However, effective inhibition of disease-implicated CAs should minimally affect the ubiquitously expressed isoforms, including CA I and II, to improve directed distribution of the inhibitors to the cancer-associated isoforms and reduce side effects. Four benzenesulfonamide-based inhibitors were synthesized by using the tail approach and displayed nanomolar affinities for several CA isoforms. The crystal structures of the inhibitors bound to a CA IX mimic and CA II are presented. Further in silico modeling was performed with the inhibitors docked into CA I and XII to identify residues that contributed to or hindered their binding interactions. These structural studies demonstrated that active-site residues lining the hydrophobic pocket, especially positions 92 and 131, dictate the positional binding and affinity of inhibitors, whereas the tail groups modulate CA isoform specificity. Geometry optimizations were performed on each ligand in the crystal structures and showed that the energetic penalties of the inhibitor conformations were negligible compared to the gains from active-site interactions. These studies further our understanding of obtaining isoform specificity when designing small molecule CA inhibitors.

Structural Study of a Flexible Active Site Loop in Human Indoleamine 2,3-Dioxygenase and Its Functional Implications.

Human indoleamine 2,3-dioxygenase catalyzes the oxidative cleavage of tryptophan to N-formyl kynurenine, the initial and rate-limiting step in the kynurenine pathway. Additionally, this enzyme has been identified as a possible target for cancer therapy. A 20-amino acid protein segment (the JK loop), which connects the J and K helices, was not resolved in the reported hIDO crystal structure. Previous studies have shown that this loop undergoes structural rearrangement upon substrate binding. In this work, we apply a combination of replica exchange molecular dynamics simulations and site-directed mutagenesis experiments to characterize the structure and dynamics of this protein region. Our simulations show that the JK loop can be divided into two regions: the first region (JK loop(C)) displays specific and well-defined conformations and is within hydrogen bonding distance of the substrate, while the second region (JK loop(N)) is highly disordered and exposed to the solvent. The peculiar flexible nature of JK loop(N) suggests that it may function as a target for post-translational modifications and/or a mediator for protein-protein interactions. In contrast, hydrogen bonding interactions are observed between the substrate and Thr379 in the highly conserved "GTGG" motif of JK loop(C), thereby anchoring JK loop(C) in a closed conformation, which secures the appropriate substrate binding mode for catalysis. Site-directed mutagenesis experiments confirm the key role of this residue, highlighting the importance of the JK loop(C) conformation in regulating the enzymatic activity. Furthermore, the existence of the partially and totally open conformations in the substrate-free form suggests a role of JK loop(C) in controlling substrate and product dynamics.