New challenges for photopharmacology and computational modeling.

Mous et al. recently reported the molecular mechanism of chloride transport through a light-activated pumping rhodopsin, a key process involved in a range of cellular functions. Their results open exciting new challenges for photopharmacology and computational modeling that should be addressed in the coming years.

Sterically driven current reversal in a molecular motor model.

Simulations can help unravel the complicated ways in which molecular structure determines function. Here, we use molecular simulations to show how slight alterations of a molecular motor's structure can cause the motor's typical dynamical behavior to reverse directions. Inspired by autonomous synthetic catenane motors, we study the molecular dynamics of a minimal motor model, consisting of a shuttling ring that moves along a track containing interspersed binding sites and catalytic sites. The binding sites attract the shuttling ring while the catalytic sites speed up a reaction between molecular species, which can be thought of as fuel and waste. When that fuel and waste are held in nonequilibrium steady-state concentrations, the free energy from the reaction drives directed motion of the shuttling ring along the track. Using this model and nonequilibrium molecular dynamics, we show that the shuttling ring's direction can be reversed by simply adjusting the spacing between binding and catalytic sites on the track. We present a steric mechanism behind the current reversal, supported by kinetic measurements from the simulations. These results demonstrate how molecular simulation can guide future development of artificial molecular motors.

The pseudotorsional space of RNA.

The characterization of the conformational landscape of the RNA backbone is rather complex due to the ability of RNA to assume a large variety of conformations. These backbone conformations can be depicted by pseudotorsional angles linking RNA backbone atoms, from which Ramachandran-like plots can be built. We explore here different definitions of these pseudotorsional angles, finding that the most accurate ones are the traditional η (eta) and θ (theta) angles, which represent the relative position of RNA backbone atoms P and C4'. We explore the distribution of η - θ in known experimental structures, comparing the pseudotorsional space generated with structures determined exclusively by one experimental technique. We found that the complete picture only appears when combining data from different sources. The maps provide a quite comprehensive representation of the RNA accessible space, which can be used in RNA-structural predictions. Finally, our results highlight that protein interactions lead to significant changes in the population of the η - θ space, pointing toward the role of induced-fit mechanisms in protein-RNA recognition.

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.

The emergence of machine learning force fields in drug design.

In the field of molecular simulation for drug design, traditional molecular mechanic force fields and quantum chemical theories have been instrumental but limited in terms of scalability and computational efficiency. To overcome these limitations, machine learning force fields (MLFFs) have emerged as a powerful tool capable of balancing accuracy with efficiency. MLFFs rely on the relationship between molecular structures and potential energy, bypassing the need for a preconceived notion of interaction representations. Their accuracy depends on the machine learning models used, and the quality and volume of training data sets. With recent advances in equivariant neural networks and high-quality datasets, MLFFs have significantly improved their performance. This review explores MLFFs, emphasizing their potential in drug design. It elucidates MLFF principles, provides development and validation guidelines, and highlights successful MLFF implementations. It also addresses potential challenges in developing and applying MLFFs. The review concludes by illuminating the path ahead for MLFFs, outlining the challenges to be overcome and the opportunities to be harnessed. This inspires researchers to embrace MLFFs in their investigations as a new tool to perform molecular simulations in drug design.

Engineering of substrate specificity in a plant cell-wall modifying enzyme through alterations of carboxyl-terminal amino acid residues.

Structural determinants of substrate recognition remain inadequately defined in broad specific cell-wall modifying enzymes, termed xyloglucan xyloglucosyl transferases (XETs). Here, we investigate the Tropaeolum majus seed TmXET6.3 isoform, a member of the GH16_20 subfamily of the GH16 network. This enzyme recognises xyloglucan (XG)-derived donors and acceptors, and a wide spectrum of other chiefly saccharide substrates, although it lacks the activity with homogalacturonan (pectin) fragments. We focus on defining the functionality of carboxyl-terminal residues in TmXET6.3, which extend acceptor binding regions in the GH16_20 subfamily but are absent in the related GH16_21 subfamily. Site-directed mutagenesis using double to quintuple mutants in the carboxyl-terminal region - substitutions emulated on barley XETs recognising the XG/penta-galacturonide acceptor substrate pair - demonstrated that this activity could be gained in TmXET6.3. We demonstrate the roles of semi-conserved Arg238 and Lys237 residues, introducing a net positive charge in the carboxyl-terminal region (which complements a negative charge of the acidic penta-galacturonide) for the transfer of xyloglucan fragments. Experimental data, supported by molecular modelling of TmXET6.3 with the XG oligosaccharide donor and penta-galacturonide acceptor substrates, indicated that they could be accommodated in the active site. Our findings support the conclusion on the significance of positively charged residues at the carboxyl terminus of TmXET6.3 and suggest that a broad specificity could be engineered via modifications of an acceptor binding site. The definition of substrate specificity in XETs should prove invaluable for defining the structure, dynamics, and function of plant cell walls, and their metabolism; these data could be applicable in various biotechnologies.

Pyrolytic conversion of glucose into hydroxymethylfurfural and furfural: Benchmark quantum-chemical calculations.

Quantum chemical methods have been intensively applied to study the pyrolytic conversion of glucose into hydroxymethylfurfural (HMF) and furfural (FF). Herein, we collect the most relevant mechanistic proposals from the recent literature and organize them into a single reaction network. All the transition structures (TSs) and intermediates are characterized using highly accurate ab initio methods and the possible reaction pathways are assessed in terms of the Gibbs energies of the TSs and intermediates with respect to ß-glucopyranose, selecting a 2D ideal-gas standard state at 773 K to represent the pyrolysis conditions. Several pathways can lead to the formation of both HMF and FF passing through rate-determining TSs that have ΔG‡ values of ~49-50 kcal/mol. Both water-assisted mechanisms and nonspecific environmental effects have a minor impact on the Gibbs energy profiles. We find that the HMF → FF + CH2O fragmentation has a small ΔrxnG value and an accessible ΔG‡ barrier. Our computational results, which are in consonance with the kinetic parameters derived from lumped models, the results of isotopic labeling experiments and the reported HMF/FF molecular ratios, could be useful for modeling studies including on nonequilibrium kinetic effects that may render more information about product yields and the relevance of the various pathways.

Fast vibrational analysis of molecular systems.

The development of infrared difference spectroscopy provides unprecedented insights on structures of complex molecules like metalloproteins. However, the relevant information can be hard to find among the many bands of the vibrational spectra. The ab initio modeling is very helpful to assign the frequencies to vibrational modes but it is a challenge to process the huge quantity of data into descriptors useful for experimentalists. To this end, we developed a new tool called VIBMOL allowing to analyze vibrational modes of molecules from hessian matrices calculated with common quantum chemistry codes. VIBMOL program runs on Unix machines. Through a new graphical interface, the users can calculate the normal modes of molecules, visualize them, simulate infrared spectra, and explore the Potential Energy Distribution of normal modes among any set of vibration coordinates. It is combined with an interface program (gosdmu) formatting relevant data from the GAUSSIAN program. VIBMOL code is available upon request to the authors. A discussion is provided to help the readers to choose between a large choice of different software and it shows how VIBMOL can make the IR assignment easier in the context of collaborations with experimentalists.

An influence of electronic structure theory method, thermodynamic and implicit solvation corrections on the organic carbonates conformational and binding energies.

An impact of an electronic structure or force field method, gas-phase thermodynamic correction, and continuum solvation model on organic carbonate clusters (S)n conformational and binding energies is explored. None of the tested force field (GFN-FF, GAFF, MMFF94) and standard semiempirical methods (PM3, AM1, RM1, PM6, PM6-D3, PM6-D3H4, PM7) can reproduce reference RI-SCS-MP2 conformational energies. Tight-binding GFNn-xTB methods provide more realistic conformational energies which are accurate enough to discard the least stable conformers. The effect of thermodynamic correction is moderate and can be ignored if the gas phase conformational stability ranking is a goal. The influence of continuum solvation is stronger, especially if reinforced with the Gibbs free energy thermodynamic correction, and results in the reduced spread of conformational energies. The cluster formation binding energies strongly depend on a particular approach to vibrational thermochemistry with the difference between traditional harmonic and modified scaled rigid - harmonic oscillator approximations reaching 10 kcal mol-1.

Ichor: A Python library for computational chemistry data management and machine learning force field development.

We present ichor, an open-source Python library that simplifies data management in computational chemistry and streamlines machine learning force field development. Ichor implements many easily extensible file management tools, in addition to a lazy file reading system, allowing efficient management of hundreds of thousands of computational chemistry files. Data from calculations can be readily stored into databases for easy sharing and post-processing. Raw data can be directly processed by ichor to create machine learning-ready datasets. In addition to powerful data-related capabilities, ichor provides interfaces to popular workload management software employed by High Performance Computing clusters, making for effortless submission of thousands of separate calculations with only a single line of Python code. Furthermore, a simple-to-use command line interface has been implemented through a series of menu systems to further increase accessibility and efficiency of common important ichor tasks. Finally, ichor implements general tools for visualization and analysis of datasets and tools for measuring machine-learning model quality both on test set data and in simulations. With the current functionalities, ichor can serve as an end-to-end data procurement, data management, and analysis solution for machine-learning force-field development.

Implementation of machine learning protocols to predict the hydrolysis reaction properties of organophosphorus substrates using descriptors of electron density topology.

Prediction of catalytic reaction efficiency is one of the most intriguing and challenging applications of machine learning (ML) algorithms in chemistry. In this study, we demonstrated a strategy for utilizing ML protocols applied to Quantum Theory of Atoms In Molecules (QTAIM) parameters to predict the ability of the A17 L47K catalytic antibody to covalently capture organophosphate pesticides. We found that the novel "composite" DFT functional B97-3c could be effectively employed for fast and accurate initial geometry optimization, aligning well with the input dataset creation. QTAIM descriptors proved to be well-established in describing the examined dataset using density-based and hierarchical clustering algorithms. The obtained clusters exhibited correlations with the chemical classes of the input compounds. The precise physical interpretation of the QTAIM properties simplifies the explanation of feature impact for both supervised and unsupervised ML protocols. It also enables acceleration in the search for entries with desired properties within large databases. Furthermore, our findings indicated that Ridge Regression with Laplacian kernel and CatBoost Regressor algorithms demonstrated suitable performance in handling small datasets with non-trivial dependencies. They were able to predict the actual reaction barrier values with a high level of accuracy. Additionally, the CatBoost Classifier proved reliable in discriminating between "active" and "inactive" compounds.

Task-Similarity is a Crucial Factor for Few-Shot Meta-Learning of Structure-Activity Relationships.

Machine learning models support computer-aided molecular design and compound optimization. However, the initial phases of drug discovery often face a scarcity of training data for these models. Meta-learning has emerged as a potentially promising strategy, harnessing the wealth of structure-activity data available for known targets to facilitate efficient few-shot model training for the specific target of interest. In this study, we assessed the effectiveness of two different meta-learning methods, namely model-agnostic meta-learning (MAML) and adaptive deep kernel fitting (ADKF), specifically in the regression setting. We investigated how factors such as dataset size and the similarity of training tasks impact predictability. The results indicate that ADKF significantly outperformed both MAML and a single-task baseline model on the inhibition data. However, the performance of ADKF varied across different test tasks. Our findings suggest that considerable enhancements in performance can be anticipated primarily when the task of interest is similar to the tasks incorporated in the meta-learning process.

Determinants of Product Outcome in Two Sesquiterpene Synthases from the Thermotolerant Bacterium Rubrobacter radiotolerans.

Rubrobacter radiotolerans nerolidol synthase (NerS) and trans-α-bergamotene synthase (BerS) are among the first terpene synthases (TSs) discovered from thermotolerant bacteria, and, despite sharing the same substrate, make terpenoid products with different carbon scaffolds. Here, the potential thermostability of NerS and BerS was investigated, and NerS was found to retain activity up to 55 °C. A library of 22 NerS and BerS variants was designed to probe the differing reaction mechanisms of NerS and BerS, including residues putatively involved in substrate sequestration, cation-π stabilisation of reactive intermediates, and shaping of the active site contour. Two BerS variants showed improved in vivo titres vs the WT enzyme, and also yielded different ratios of the related sesquiterpenoids (E)-ß-farnesene and trans-α-bergamotene. BerS-L86F was proposed to encourage substrate isomerisation by cation-π stabilisation of the first cationic intermediate, resulting in a greater proportion of trans-α-bergamotene. By contrast, BerS-S82L significantly preferred (E)-ß-farnesene formation, attributed to steric blocking of the isomerisation step, consistent with what has been observed in several plant TSs. Our work highlights the importance of isomerisation as a key determinant of product outcome in TSs, and shows how a combined computational and experimental approach can characterise TSs and variants with improved and altered functionality.

Gas-phase structure of polymer ions: Tying together theoretical approaches and ion mobility spectrometry.

An increasing number of studies take advantage of ion mobility spectrometry (IMS) coupled to mass spectrometry (IMS-MS) to investigate the spatial structure of gaseous ions. Synthetic polymers occupy a unique place in the field of IMS-MS. Indeed, due to their intrinsic dispersity, they offer a broad range of homologous ions with different lengths. To help rationalize experimental data, various theoretical approaches have been described. First, the study of trend lines is proposed to derive physicochemical and structural parameters. However, the evaluation of data fitting reflects the overall behavior of the ions without reflecting specific information on their conformation. Atomistic simulations constitute another approach that provide accurate information about the ion shape. The overall scope of this review is dedicated to the synergy between IMS-MS and theoretical approaches, including computational chemistry, demonstrating the essential role they play to fully understand/interpret IMS-MS data.

Scaffold-Oriented Asymmetric Catalysis: Conformational Modulation of Transition State Multivalency during a Catalyst-Controlled Assembly of a Pharmaceutically Relevant Atropisomer.

A new class of superbasic, bifunctional peptidyl guanidine catalysts is presented, which enables the organocatalytic, atroposelective synthesis of axially chiral quinazolinediones. Computational modeling unveiled the conformational modulation of the catalyst by a novel phenyl urea N-cap, that preorganizes the structure into the active, folded state. A previously unanticipated noncovalent interaction involving a difluoroacetamide acting as a hybrid mono- or bidentate hydrogen bond donor emerged as a decisive control element inducing atroposelectivity. These discoveries spurred from a scaffold-oriented project inspired from a fascinating investigational BTK inhibitor featuring two stable chiral axes and relies on a mechanistic framework that was foreign to the extant lexicon of asymmetric catalysis.

Influence of Pnictogen and Ligand Framework on the Lewis Acidity and Steric Environment in Pnictogen Pincer Complexes.

Pnictogen pincer complexes are a fascinating class of compounds due to their dynamic molecular and electronic structures, and valuable stoichiometric or catalytic reactivity. As recognition of their unique chemistry has grown, so too has the library of pincer ligands employed and pnictogen centres engaged to prepare them. Here we computationally study how the choice of pincer ligand framework and pnictogen influence the electronic and steric outcomes of the complexes obtained. The most relevant electronic parameter is the pnictogen-centred electrophilicity, which has been quantified by fluoride ion affinities and LUMO energies, while the most relevant steric parameter is the crowding around the central pnictogen, which has been quantified by the %Vbur values and visualized using steric maps. The resulting trends are analyzed with reference to binding pocket size, acceptor orbital type, electronic delocalization, π-donor strengths, and heteroatom incorporation. Thus, considering 16 ligand frameworks and 4 heavy pnictogen centres, this study provides a broad-spectrum view of stereo-electronic variation in pnictogen pincer complexes, which, together with a recent study on geometric variation in the same family, provides a substantial dataset to guide future molecular design and reactivity studies.

Unraveling the Source of Self-Induced Diastereomeric Anisochronism in Chiral Dipeptides.

Mastering of analytical methods for accurate quantitative determinations of enantiomeric excess is a crucial aspect in asymmetric catalysis, chiral synthesis, and pharmaceutical applications. In this context, the phenomenon of Self-Induced Diastereomeric Anisochronism (SIDA) can be exploited in NMR spectroscopy for accurate determinations of enantiomeric composition, without using a chiral auxiliary that could interfere with the spectroscopic investigation. This phenomenon can be particularly useful for improving the quantitative analysis of mixtures with low enantiomeric excesses, where direct integration of signals can be tricky. Here, we describe a novel analysis protocol to correctly determine the enantiomeric composition of scalemic mixtures and investigate the thermodynamic and stereochemical features at the basis of SIDA. Dipeptide derivatives were chosen as substrates for this study, given their central role in drug design. By integrating the experiments with a conformational stochastic search that includes entropic contributions, we provide valuable information on the dimerization thermodynamics, the nature of non-covalent interactions leading to self-association, and the differences in the chemical environment responsible for the anisochronism, highlighting the importance of different stereochemical arrangement and tight association for the distinction between homochiral and heterochiral adducts. An important role played by the counterion was pointed out by computational studies.

Prediction of Covalent Metal-Metal Bonding in Cp-M-M'-Nacnac Complexes of Group 2 and 12 Metals (Be, Mg, Ca, Zn, Cd, Hg).

Bimetallic CpMM'Nacnac molecules with group 2 and 12 metals (M=Be, Mg, Ca, Zn, Cd, Hg) that contain novel metal-metal bonding have been investigated in a theoretical study of their molecular and electronic structure, thermodynamic stability, and metal-metal bonding. In all cases the metal-metal bonds are characterized as electron-sharing covalent single bonds from natural bond orbital (NBO) and energy-decomposition analysis with natural orbitals of chemical valence (EDA-NOCV) analysis. The sum of [MM'] charges is relatively constant, with all complexes exhibiting a [MM']2+ core. Quantum theory of atoms in molecules (QTAIM) analysis indicates the presence of non-nuclear attractors (NNA) in the metal-metal bonds of the BeBe, MgMg, and CaCa complexes. There is substantial electron density (0.75-1.33 e) associated with the NNAs, which indicates that these metal-metal bonds, while classified as covalent electron-sharing bonds, retain significant metallic character that can be associated with reducing reactivity of the complex. The predicted stability of these complexes, combined with their novel covalent metal-metal bonding and potential as reducing agents, make them appealing targets for the synthesis of new metal-metal bonds.

Migration of para-Nitrophenyl Groups in Methyl Pyranosides: Configuration and Conformation Determine the Kinetics.

para-Nitrophenyl (PNP) ethers of glycosides are important building blocks en route to functional carbohydrates. They are stable in neutral media, however, under basic conditions such as during the Zemplén deacylation of sugars, aryl migration is frequently observed. We have employed a library of O-PNP-substituted methyl glycosides of the manno-, galacto-, gluco- and altro-series to study the kinetics of aryl migration in MeOH/sodium methoxide using NMR spectroscopy revealing that migration between cis-oriented OH groups is faster than between trans-oriented ones. The rate constants of migration decrease in the order of Alt>Man>Gal>Glc and are related to the energy barriers of chair conformation inversion. The energy profile of the 3 to 4-PNP migration in methyl mannoside was calculated using DFT methods suggesting the Meisenheimer complex is an intermediate of PNP migration and that coordination of the sodium cation has a major impact on the energy profile.

Antiaromaticity of Cycloheptatrienyl Anions: Structure, Acidity, and Magnetic Properties.

Investigations of the nature and degree of antiaromaticity of cycloheptatrienyl anion derivatives using both experimental and computational tools are presented. The ground state of cycloheptatrienyl anion in the gas phase is triplet, planar and Baird-aromatic. In DMSO, it assumes a singlet distorted allylic form with a paratropic ring current. The other derivatives in both phases assume either allylic or diallylic conformations depending on the substituent pattern. A combination of experimental and computational methods was used to determine the pKa values of 16 derivatives in DMSO, which ranged from 36 to -10.7. We revealed that the stronger stabilization of the anionic system, which correlates with acidity, does not necessarily imply a lower degree of antiaromaticity in terms of magnetic properties. Conversely, the substitution pattern first affects the geometry of the ring through the bulkiness of the substituents and their better conjugation with a more distorted system. Consequently, the distortion reduces the cyclic conjugation in the π-system and thereby decreases the paratropic current in a magnetic field, which manifests itself as a decrease in the NICS. The triplet-state geometries and magnetic properties are nearly independent on the substitution pattern, which is typical for simple aromatic systems.