QM/MM methods for biomolecular systems

Computational studies for the development of extracellular vesicle-based biosensors

Cancer affects millions of people, and early detection and efficient treatment are two strong levers to hurdle this disease. Recent studies on exosomes, a subset of extracellular vesicles, have deliberately shown the potential to function as a biomarker or treatment tool, thereby attracting the attention of researchers who work on developing biosensors. Due to the ability of computational methods to predict of the behavior of biomolecules, the combination of experimental and computational methods would enhance the analytical performance of the biosensor, including sensitivity, accuracy, and specificity, even detecting such vesicles from bodily fluids. In this regard, the role of computational methods such as molecular docking, molecular dynamics simulation, and density functional theory is overviewed in the development of biosensors. This review highlights the investigations and studies that have been reported using these methods to design exosome-based biosensors. This review concludes with the role of the quantum mechanics/molecular mechanics method in the investigation of chemical processes of biomolecular systems and the deficiencies in using this approach to develop exosome-based biosensors. In addition, the artificial intelligence theory is explained briefly to show its importance in the study of these biosensors.

Molecular surfaces modeling: Advancements in deep learning for molecular interactions and predictions

Molecular surface analysis can provide a high-dimensional, rich representation of molecular properties and interactions, which is crucial for enabling powerful predictive modeling and rational molecular design across diverse scientific and technological domains. With remarkable successes achieved by artificial intelligence (AI) in different fields such as computer vision and natural language processing, there is a growing imperative to harness AI's potential in accelerating molecular discovery and innovation. The integration of AI techniques with molecular surface analysis has opened up new frontiers, allowing researchers to uncover hidden patterns, relationships, and design principles that were previously elusive. By leveraging the complementary strengths of molecular surface representations and advanced AI algorithms, scientists can now explore chemical space more efficiently, optimize molecular properties with greater precision, and drive transformative advancements in areas like drug development, materials engineering, and catalysis. In this review, we aim to provide an overview of recent advancements in the field of molecular surface analysis and its integration with AI techniques. These AI-driven approaches have led to significant advancements in various downstream tasks, including interface site prediction, protein-protein interaction prediction, surface-centric molecular generation and design.

Systematic molecular profiling of non-native N6-substitution effects on m6A binding to the YTH domains of human RNA m6A readers in diabetes

The RNA N6-adenosine methylation, resulting in N6-methyl adenosine (m6A), is one of the most important post-transcriptional modification events in the eukaryotic transcriptome, which is dynamically regulated by methyltransferases (writers), recognition proteins (readers) and demethylases (erasers). Human has five m6A readers namely YTHDC1, YTHDC2, YTHDF1, YTHDF2 and YTHDF3 that specifically recognize and bind to the methylated m6A residue of RNA through their YT521-B homology (YTH) domains, which have been involved in the pathogenesis of diabetes mellitus and its diverse complications such as diabetic nephropathy. Instead of the native N6-methylation, we herein attempted to explore the molecular effect of various non-native N6-substitutions on adenosine (A) binding behavior to YTH domains. A systematic interaction profile of 40 reported N6-substituted adenosine (x6A) mononucleotides with 5 human reader YTH domains was created computationally. Heuristic clustering of the profile divided these YTH domains and these x6A mononucleotides into two subfamilies and three classes, respectively; they represent distinct intrinsic interaction modes between the domains and mononucleotides. Statistical survey unraveled that the volume (Vg) and hydrophobicity (Hg) of N6-substituted chemical groups exhibit linear and nonlinear correlations with the binding energy (ΔGttl) of x6A mononucleotides to YTH domains, respectively; N6-substitutions with moderate size and weak polarity are favorable for the x6A binding. From the profile the N6-bromomethyl adenosine (brm6A) was identified as a potent binder of YTHDF2 YTH domain; its affinity was improved significantly by 77.2-fold from A and considerably by 19.5-fold from m6A. Structural modeling observed that the N6-bromomethyl group of brm6A is tightly packed against an aromatic cage defined by the Trp432-Trp486-Trp491 triad of YTHDF2 YTH domain. Electron-correlation analysis revealed that the bromine atom can form geometrically and energetically satisfactory halogen-π interactions with the aromatic cage, thus conferring considerable affinity and specificity to the domain-brm6A interaction.

Assessment of Free Energies From Electrostatic Embedding Density Functional Tight Binding-Based/Molecular Mechanics in Periodic Boundary Conditions

Electrostatic embedding quantum mechanics/molecular mechanics (QM/MM) methods in periodic boundary conditions (PBC) can successfully describe the condensed phase reactivity of a fragment treated at the QM level with an atomistic description of an electrostatic environment treated at the MM level. The computational cost of ab initio QM methods limits the phase space sampling, thus affecting statistical quantities like free energies. Here, we describe the implementation of a PBC-adapted QM/MM model based on the semi-empirical density-functional based tight-binding (DFTB) method within the GAMESS-US quantum package interfaced with Tinker. Further, we take advantage of the free energy methods provided by a newly developed interface with the PLUMED plugin. The versatility of the implementation is illustrated by the prediction of the free energy profile for three different families of reactions in solution. Overall, using the DFTB/MM, it has been possible to obtain results that are at least in a qualitatively agreement with respect to the experimental data or high-level ab initio simulations.

The Amsterdam Modeling Suite

In this paper, we present the Amsterdam Modeling Suite (AMS), a comprehensive software platform designed to support advanced molecular and materials simulations across a wide range of chemical and physical systems. AMS integrates cutting-edge quantum chemical methods, including Density Functional Theory (DFT) and time-dependent DFT, with molecular mechanics, fluid thermodynamics, machine learning techniques, and more, to enable multi-scale modeling of complex chemical systems. Its design philosophy allows for seamless coupling between components, facilitating simulations that range from small molecules to complex biomolecular and solid-state systems, making it a versatile tool for tackling interdisciplinary challenges, both in industry and in academia. The suite also emphasizes user accessibility, with an intuitive graphical interface, extensive scripting capabilities, and compatibility with high-performance computing environments.

Unraveling the complexity of organophosphorus pesticides: Ecological risks, biochemical pathways and the promise of machine learning

Organophosphorus pesticides (OPPs) are widely used in agriculture but pose significant ecological and human health risks due to their persistence and toxicity in the environment. While microbial degradation offers a promising solution, gaps remain in understanding the enzymatic mechanisms, degradation pathways, and ecological impacts of OPP transformation products. This review aims to bridge these gaps by integrating traditional microbial degradation research with emerging machine learning (ML) technologies. We hypothesize that ML can enhance OPP degradation studies by improving the efficiency of enzyme discovery, pathway prediction, and ecological risk assessment. Through a comprehensive analysis of microbial degradation mechanisms, environmental factors, and ML applications, we propose a novel framework that combines biochemical insights with data-driven approaches. Our review highlights the potential of ML to optimize microbial strain screening, predict degradation pathways, and identify key active sites, offering innovative strategies for sustainable pesticide management. By integrating traditional research with cutting-edge ML technologies, this work contributes to the journal's scope by promoting eco-friendly solutions for environmental protection and pesticide pollution control.

High-performance QM/MM Enhanced Sampling Molecular Dynamics Simulations with GENESIS SPDYN and QSimulate-QM

A new module for quantum mechanical/molecular mechanical (QM/MM) calculations is implemented in a molecular dynamics (MD) program, SPDYN in GENESIS, interfaced with an electronic structure program, QSimulate-QM. The periodic boundary condition (PBC) in QM/MM simulation is incorporated via QM calculation in real space with duplicated MM charges and particle mesh Ewald (PME) calculation with QM and MM charges. A highly optimized code is implemented in QSimulate-QM, particularly for the density functional tight-binding (DFTB) method, where the interaction between the QM and MM regions is computed utilizing multipole expansions. Together with highly parallelized algorithms in SPDYN, the developed program performs MD simulations based on DFTB in the QM size of ∼100 atoms and MM of ∼100,000 atoms with a better performance than 1 ns/day using one computer node. This feature paves the way for QM/MM-MD enhanced sampling simulations. Various enhanced sampling methods in GENESIS, namely, generalized replica exchange solute tempering (gREST), replica-exchange umbrella sampling (REUS), and path sampling with the string method, are demonstrated at the QM/MM level to compute the Ramachandran plot of alanine dipeptide and the potential of mean force (PMF) of a proton transfer reaction in an enzyme.

Reaction Mechanism Path Sampling Based on Parallel Cascade Selection QM/MM Molecular Dynamics Simulation: PaCS-Q

Quantum mechanics/molecular mechanics (QM/MM) molecular dynamics (MD) simulations are essential for elucidating complex biochemical reaction mechanisms. However, conventional enhanced sampling methods, such as umbrella sampling and metadynamics, often face limitations in computational cost, sampling completeness, and reliance on predefined reaction coordinates. To address these challenges, we developed Parallel Cascade Selection QM/MM MD (PaCS-Q) simulation, a novel strategy that efficiently explores reaction pathways by iteratively identifying high-potential structures for configurational transitions without predefined biases or external constraints. PaCS-Q directly tracks changes in bond distances over time, enabling the identification of transition states and intermediates. Validation of the Claisen rearrangement in chorismate mutase and the peptidyl aldehyde reaction in the Zika virus NS2B/NS3 serine protease demonstrated accurate pathway capture, reduced computational costs, and efficient sampling. With its user-friendly workflow, PaCS-Q broadens accessibility for computational and experimental researchers, offering a robust tool for studying enzymatic mechanisms with high accuracy and efficiency.

Exact Two-Component Relativistic Polarizable Density Embedding

We have implemented the fragment-based polarizable density embedding (PDE) model within a relativistic framework building on the eXact 2-Component (X2C) relativistic Hamiltonian, thereby taking the PDE method to a relativistic framework. The PDE model provides a robust solution to the electron-leakage problem, and we show that this newly implemented model offers an accurate way to model solvated systems possessing significant relativistic effects. To demonstrate the model's performance, we perform comparative calculations of the K- and L2,3-edge spectra of water-solvated cysteine (both protonated and deprotonated) with the X2C Hamiltonian. Particularly, with counterions such as Na+ in the solvent, electron leakage clearly shows in the older polarizable embedding model through spurious peaks in the spectra. However, when the PDE model is employed, these spurious peaks disappear.

Probing the Effect of Protein and Inhibitor Conformational Flexibility on the Reaction of Rocelitinib-Like Covalent Inhibitors of Epidermal Growth Factor Receptor. A Quantum Mechanics/Molecular Mechanics Study

Epidermal growth factor receptor (EGFR) is a tyrosine kinase and a validated target for non-small cell lung cancer (NSCLC). Drug discovery efforts on this target initially focused on traditional competitive, reversible ATP-binding site inhibitors; however, irreversible covalent binding EGFR inhibitors have become increasingly more popular. Covalent EGFR inhibitors have been developed using a range of different scaffolds, and unsurprisingly, the incorporation of an electrophilic acrylamide group can result in sizable orientation differences relative to the Cys797 nucleophile and the Asp800 general base. In this work, we report a QM/MM study aiming to better understand the aspects of covalent adduct formation, including the role of protein flexibility on chemical reactivity, the impact of electrophile location within the ATP binding site, and the impact of the acrylamide conformation (s-cis vs s-trans). We focus here on the diaminopyrimidine scaffold, as exemplified by Rocelitinib, where the electrophile is attached to its back pocket binding group. Our goal is to elucidate how electrophilic groups can be incorporated onto different inhibitor scaffolds targeting reactive active site residues. We find that irrespective of the EGFR MD conformation chosen, acrylamide, in both the s-cis or s-trans, can undergo reaction with rate-determining barriers of ∼20 kcal/mol. Interestingly, the nature of the rate-determining step for Rocelitinib-like inhibitors was found to be either direct nucleophilic attack or keto-enol tautomerization, depending on the precise protein and inhibitor conformation.

Molecular dynamics study on the mitigation of radiation damage caused by electron pulses

The reduction of radiation damage represents a long-term objective for electron microscopists, particularly those engaged in the study of biological and organic matter. Recently, electron pulses in ultrafast transmission electron microscopy have been demonstrated to serve as a damage mitigation technique for radiation-sensitive materials. Nevertheless, the underlying mechanism of the mitigation effects remains unclear. In this study, we investigate the radiation damage of graphene induced by pulsed electrons using molecular dynamics simulations within the framework of binary elastic collisions. For electron irradiation at 200 keV, it was found that the pulsed electron beam corresponds to a larger threshold angle (1.4 rad) than that for a random beam (1.0 rad). This is because two electrons can be prevented from briefly interacting with the same or a neighboring atom by the use of well-controlled electron pulses. While such a mitigation of radiation damage is only apparent near the threshold angle, and there are likely other reduction mechanisms, our results provide insight into the mitigated radiation damage of electron pulses.

Multiscale Modeling and Simulation of Zwitterionic Anti-fouling Materials

Zwitterionic materials with cationic and anionic moieties in the same chain, being electrically neutral, have excellent hydrophilicity, stability, biocompatibility, and outstanding anti-biofouling performance. Because of their unique properties, zwitterionic materials are widely applied to membrane separation, drug delivery, surface coating, etc. However, what is the root of their unique properties? It is necessary to study the structure-property relationships of zwitterionic compounds to guide the design and development of zwitterionic materials. Modeling and simulation methods are considered to be efficient technologies for understanding advanced materials in principle. This Review systematically summarizes the computational exploration of zwitterionic materials in recent years. First, the classes of zwitterionic materials are summarized. Second, the different scale simulation methods are introduced briefly. To reveal the structure-property relationships of zwitterionic materials, multiscale modeling and simulation studies at different spatial and temporal scales are summarized. The study results indicated that the strong electrostatic interaction between zwitterions with water molecules promotes formation of a stable hydration layer, namely, superhydrophilicity, leading to the excellent anti-fouling properties. Finally, we offer our viewpoint on the development and application of simulation techniques on zwitterionic materials exploration in the future. This work establishes a bridge from atomic and molecular scales to mesoscopic and macroscopic scales and helps to provide an in-depth understanding of the structure-property relationships of zwitterionic materials.

Modeling Raman Spectra in Complex Environments: From Solutions to Surface-Enhanced Raman Scattering

This perspective highlights the essential physicochemical factors required for accurate computational modeling of Raman and Resonance Raman signals in complex environments. It highlights the theoretical challenges for obtaining a balanced quantum mechanical description of the molecular target, integration of target-environment interactions into the Hamiltonian, and explicit treatment of strong interactions such as hydrogen bonding. The dynamical sampling of solute-solvent phase space and the incorporation of plasmonic effects for Surface-Enhanced Raman Scattering (SERS) are also addressed. Through selected applications, we illustrate how these factors influence Raman signals and propose a framework to tackle these challenges effectively, advancing the reliability of theoretical Raman spectroscopy in real-world scenarios.

Efficient Computational Strategies of the Cluster-in-Molecule Local Correlation Approach for Interaction Energies of Large Host-Guest Systems

We propose a heterogeneously accelerated reduced cluster-in-molecule (CIM) local correlation approach for calculating host-guest interaction energies. The essence of this method is to compute only the clusters that make significant contributions to the interaction energies while approximately neglecting those clusters with smaller contributions. Benchmark calculations at the CIM resolution-of-identity second-order Mo̷ller-Plesset perturbation (CIM-RI-MP2) or CIM spin-component-scaled RI-MP2 (CIM-SCS-RI-MP2) levels, involving three medium-sized protein-ligand structures, demonstrate that the reduced CIM method achieves over 48% time savings without compromising accuracy, as the interaction energy error remains within 0.5 kcal/mol compared to the full CIM method. To further enhance cluster computation efficiency, we developed a heterogeneous parallel version of the CIM-(SCS-)RI-MP2 method. It achieves over 93% internode parallel efficiency and over 98% multi-GPU card parallel efficiency for the tested large complexes. Ultimately, the hardware-accelerated reduced CIM-(SCS-)RI-MP2 method is applied to calculate the interaction energies of six protein-ligand systems, ranging from 913 to 1425 atoms. Remarkably, the method requires only 4.3-22.8% of the clusters to achieve accurate results, and under the condition of using only a single node, the wall time is within 2 days. Additionally, the reduced CIM domain-based local pair natural orbital coupled cluster with singles, doubles, and perturbative triples [CIM-DLPNO-CCSD(T)] method is successfully applied to the calculation of a 1425-atom protein-ligand system. These computations demonstrate the capability of a specific electronic structure to accurately calculate interaction energies for large host-guest systems.

Multiscale Force Field Model Based on a Graph Neural Network for Complex Chemical Systems

Inspired by the QM/MM methodology, the ML/MM approach introduces a new opportunity for multiscale simulation, improving the balance between accuracy and computational efficiency. Benefited from the rapid advancements in molecular embedding methods, density functional theory level quantum mechanical (QM) calculations within the QM/MM framework can be accelerated by several orders of magnitude through the application of machine learning (ML) potential energy surfaces. As a problem inherited from the QM/MM methodology, challenges exist in designing the interactions between machine learning and molecular mechanics (MM) regions. In this study, electrostatic interactions between machine learning and MM atoms are treated by using a graphical neural network based on stationary perturbation theory. In this protocol, we process coordinates and MM charges to yield electrostatic energy and forces, resulting in a high-performance electrostatic embedding ML/MM architecture. The accuracy of the ML/MM energy was validated in aqueous solutions of alanine dipeptide and allyl vinyl ether (AVE). We investigated the transferability of parameters trained from AVE in a single solvent to various other solvents, including water, methanol, dimethyl sulfoxide, toluene, ionic liquids, and water-toluene interface environments. We then established a solvent-free protocol for data set preparation. Comparison of the free energy landscapes of the Claisen rearrangement of AVE in different solvation environments showed the catalytic effect of aqueous solutions, consistent with experiments.

Efficient Training of Neural Network Potentials for Chemical and Enzymatic Reactions by Continual Learning

Machine learning (ML) methods have emerged as an efficient surrogate for high-level electronic structure theory, offering precision and computational efficiency. However, the vast conformational and chemical space remains challenging when constructing a general force field. Training data sets typically cover only a limited region of this space, resulting in poor extrapolation performance. Traditional strategies must address this problem by training models from scratch using old and new data sets. In addition, model transferability is crucial for general force field construction. Existing ML force fields, designed for closed systems with no external environmental potential, exhibit limited transferability to complex condensed phase systems such as enzymatic reactions, resulting in inferior performance and high memory costs. Our ML/MM model, based on the Taylor expansion of the electrostatic operator, showed high transferability between reactions in several simple solvents. This work extends the strategy to enzymatic reactions to explore the transferability between more complex heterogeneous environments. In addition, we also apply continual learning strategies based on memory data sets to enable autonomous and on-the-fly training on a continuous stream of new data. By combining these two methods, we can efficiently construct a force field that can be applied to chemical reactions in various environmental media.

Advancing Multiscale Molecular Modeling with Machine Learning-Derived Electrostatics

We introduce an innovative machine learning (ML)-based framework for multiscale molecular modeling in which the ML subsystem is treated as an electrostatic entity interacting with its molecular mechanics (MM) environment through classical electrostatics. The integration of ML accuracy with multiscale modeling is accomplished by leveraging the capabilities of the ANI neural networks to predict geometry-dependent atomic partial charges at the minimal basis iterative stockholder (MBIS) level, going beyond static mechanical embedding. This ML/MM approach can closely approximate state-of-the-art multiscale quantum-classical (QM/MM) methods while significantly lowering computational requirements, thereby facilitating more efficient and precise simulations in computational chemistry. The method requires no additional training beyond the initial model setup and is integrated into Amber, one of the most widely used software suites for molecular modeling, ensuring accessibility to the broader community. We validate its performance across a variety of challenging applications, including the solvation structure, vibrational spectra, torsion free energy profiles, and protein-ligand interactions, achieving excellent agreement with QM/MM benchmarks. This framework not only advances the frontiers of multiscale modeling but also showcases the potential of machine learning to achieve quantum-level accuracy with exceptional efficiency for complex chemical systems.

Decoding Drug Discovery: Exploring A-to-Z In Silico Methods for Beginners

The drug development process is a critical challenge in the pharmaceutical industry due to its time-consuming nature and the need to discover new drug potentials to address various ailments. The initial step in drug development, drug target identification, often consumes considerable time. While valid, traditional methods such as in vivo and in vitro approaches are limited in their ability to analyze vast amounts of data efficiently, leading to wasteful outcomes. To expedite and streamline drug development, an increasing reliance on computer-aided drug design (CADD) approaches has merged. These sophisticated in silico methods offer a promising avenue for efficiently identifying viable drug candidates, thus providing pharmaceutical firms with significant opportunities to uncover new prospective drug targets. The main goal of this work is to review in silico methods used in the drug development process with a focus on identifying therapeutic targets linked to specific diseases at the genetic or protein level. This article thoroughly discusses A-to-Z in silico techniques, which are essential for identifying the targets of bioactive compounds and their potential therapeutic effects. This review intends to improve drug discovery processes by illuminating the state of these cutting-edge approaches, thereby maximizing the effectiveness and duration of clinical trials for novel drug target investigation.

QM/MM study reveals novel mechanism of KRAS and KRASG12R catalyzed GTP hydrolysis

As a crucial drug target, KRAS can regulate most cellular processes involving guanosine triphosphate (GTP) hydrolysis. However, the mechanism of GTP hydrolysis has remained controversial over the past decades. Here, several different GTP hydrolysis mechanisms catalyzed by wild-type KRAS (WT-KRAS) and KRASG12R mutants were discussed via four QM/MM calculation models. Based on the computational results, a Mg2+-coordinated H2O-mediated GTP hydrolysis mechanism was proposed. In this mechanism, a Mg2+-coordinated H2O first protonates the fully deprotonated GTP, and then the Mg2+ coordinated hydroxyl anion is generated. The Pγ-O bond is formed via the SN2 attack of the second H2O on the Pγ atom of the GTP, leading to the simultaneous cleavage of the Pγ-O bond. Meanwhile, the hydroxyl anion coordinated to Mg2+ and generated in the first step acts as a proton acceptor from water. This Mg2+ coordinated H2O-involved GTP hydrolysis mechanism may also be suitable for Mg2+-catalyzed ATP hydrolysis. Furthermore, the mechanism of GTP hydrolysis catalyzed by the KRASG12R mutant, whose hydrolysis rate was approximately 40-fold slower than WT-KRAS, was also discussed. Our QM/MM calculations reveal that GTP is easily protonated by the residue R12, and the energy barrier of GTP hydrolysis catalyzed by the KRASG12R mutant is lower than the corresponding barrier for WT-KRAS. Nevertheless, molecular dynamics (MD) simulations reveal that R12, a residue characterized by significant steric hindrance, is positioned at the GTP site where the nucleophilic attack by water occurs during Pγ-O bond formation, thereby strongly impeding the approach of water molecules to GTP. As a result, the GTP hydrolysis rate catalyzed by the KRASG12R mutant was severely impaired. Uncovering the GTP hydrolysis mechanism catalyzed by the WT-KRAS and KRASG12R mutant may also give a reasonable explanation for the relationship between the KRASG12R mutation and the occurrence of cancer. We hope this finding will provide useful guidance for drug discovery that targets KRAS.

Advanced molecular modeling of proteins: Methods, breakthroughs, and future prospects

The contemporary advancements in molecular modeling of proteins have significantly enhanced our comprehension of biological processes and the functional roles of proteins on a global scale. The application of advanced methodologies, including homology modeling, molecular dynamics simulations, and quantum mechanics/molecular mechanics strategies, has empowered numerous researchers to forecast the behavior of protein macromolecules, elucidate drug-protein interactions, and develop drugs with enhanced precision. This chapter elucidates the advent of deep learning algorithms such as AlphaFold, a notable advancement that has significantly improved the precision of intricate protein structure predictions. The recent advancements have significantly enhanced the precision of protein predictions and expedited drug discovery and development processes. Integrating approaches like multi-scale modeling and hybrid methods incorporating reliable experimental data is anticipated to revolutionize and offer more significant implications for precision medicine and targeted treatments.

Neural Network Potential with Multiresolution Approach Enables Accurate Prediction of Reaction Free Energies in Solution

We present the design and implementation of a novel neural network potential (NNP) and its combination with an electrostatic embedding scheme, commonly used within the context of hybrid quantum-mechanical/molecular-mechanical (QM/MM) simulations. Substitution of a computationally expensive QM Hamiltonian by an NNP with the same accuracy largely reduces the computational cost and enables efficient sampling in prospective MD simulations, the main limitation faced by traditional QM/MM setups. The model relies on the recently introduced anisotropic message passing (AMP) formalism to compute atomic interactions and encode symmetries found in QM systems. AMP is shown to be highly efficient in terms of both data and computational costs and can be readily scaled to sample systems involving more than 350 solute and 40,000 solvent atoms for hundreds of nanoseconds using umbrella sampling. Most deviations of AMP predictions from the underlying DFT ground truth lie within chemical accuracy (4.184 kJ mol-1). The performance and broad applicability of our approach are showcased by calculating the free-energy surface of alanine dipeptide, the preferred ligation states of nickel phosphine complexes, and dissociation free energies of charged pyridine and quinoline dimers. Results with this ML/MM approach show excellent agreement with experimental data and reach chemical accuracy in most cases. In contrast, free energies calculated with static DFT calculations paired with implicit solvent models or QM/MM MD simulations using cheaper semiempirical methods show up to ten times higher deviation from the experimental ground truth and sometimes even fail to reproduce qualitative trends.

Recombination of Autodissociated Water Ions in a Nanoscale Pure Water Droplet

The recombination of water ions has diverse scientific and practical implications, ranging from acid-base chemistry and biological systems to planetary environments and applications in fuel cell and carbon conversion technologies. While spatial confinement affects the physicochemical properties of water dynamics, its impact on the recombination process has rarely been studied. In this work, we investigate the dynamics of water, the water ion distribution, and the ion recombination process in water droplets as a function of droplet size through molecular dynamics simulations and adaptive quantum mechanical/molecular mechanical calculations. We compare the dynamics of recombination in water droplet sizes ranging from 100 to 18 000 waters, both in their interiors and on their surfaces. We found that the self-diffusion of water dramatically decreases in droplets with a diameter below 2.2 nm. Using a classical RexPoN force-field, we found that the ions in 1000 H2O's spend almost 50% of the time on the surface and 0.5 nm beneath it with a slight preference for OH- ion to reside longer on the surface. We estimate that, on average, recombination in these drops occurs at 400 ps in 1000 H2O's and 1 ns in 3000 H2O's. We also found that recombination is not limited by the local structure of the surface or the size of the droplet but can be influenced by the geometry of the water wire connecting the ions as they approach each other, which can often prevent recombination. Our results provide insights to the reaction microenvironments presented by nanoscopic water droplets.

Revisiting Oxygen Transport Features of Hemocyanin with NEVPT2 Level QM/MM Calculations

This study explores the oxygen-binding mechanism and the potential peroxo-to-bis-μ-oxo isomerization in hemocyanin (Hc) using a quantum mechanics/molecular mechanics (QM/MM) approach at the multireference NEVPT2 level of theory (QM[NEVPT2]/MM). Our results support the previously proposed mechanism for Hc oxygen binding, involving two nearly simultaneous electron-transfer (ET) steps and a triplet-singlet intersystem crossing (ISC). However, we find that the first ET step occurs prior to ISC, resulting in the formation of a stable singlet superoxide intermediate through a low-energy barrier. The second ET leads to the formation of a singlet oxy-hemocyanin species featuring the characteristic peroxo-Cu2O2 "butterfly" core. Moreover, QM[NEVPT2]/MM simulations reveal a lower-energy barrier for the peroxo-to-bis-μ-oxo isomerization compared with density functional theory (DFT), although the peroxo form remains energetically favored within the protein environment. These findings offer new insights into the behavior of the hemocyanin active site, highlighting the importance of considering both the electronic correlation and the protein environment in accurately modeling copper-oxygen interactions in biological systems.

Toward Quantum Chemical Accuracy in Absolute Protein-Ligand Binding Free Energy Calculation via Quantum Fragment Method

Accurate computation of protein-ligand binding free energy remains an elusive goal due to inherent difficulties involved in the accurate calculation of gas-phase protein-ligand interaction energy, the entropy, and the solvation energy. In this study, we explore the use of fragment quantum chemical calculations for improved accuracy in protein-ligand binding free energy calculations. The present work demonstrated that the gas-phase protein-ligand interaction energies can be accurately calculated by the molecular fractionation with conjugate caps method as verified by comparison with the full quantum calculations for several protein-ligand systems. The m06-2x/6-31+G* level of density functional theory calculation with basis set superposition error correction is found to give excellent protein-ligand interaction energies. The quantum calculated protein-ligand interaction energies are then combined with implicit solvation methods to obtain absolute binding free energies and the results are shown to be sensitive to the specific solvation models used. In particular, the accuracy of the quantum calculated binding free energies is significantly improved over that of the force field calculations using the same solvation models in terms of mean absolute errors. However, the correlation coefficients with respect to the experimental data do not show improvement over the corresponding result computed from the force field. Such result and the related analysis underscore the critical importance of solvation energies to the binding free energies and the need for developing new methods to calculate solvation energies more accurately in the future.

Mechanism of Regio- and Enantioselective Hydroxylation of Arachidonic Acid Catalyzed by Human CYP2E1: A Combined Molecular Dynamics and Quantum Mechanics/Molecular Mechanics Study

Regio- and enantioselective hydroxylation of free fatty acids by human cytochrome P450 2E1 (CYP2E1) plays an important role in metabolic regulation and has significant pathological implications. Despite extensive research, the detailed hydroxylation mechanism of CYP2E1 remains incompletely understood. To clarify the origins of regioselectivity and enantioselectivity observed for CYP2E1-mediated fatty acid hydroxylation, molecular dynamics (MD) simulations and quantum mechanics/molecular mechanics (QM/MM) calculations were performed. MD simulations provided key insights into the proximity of arachidonic acid's carbon atoms to the reactive iron(IV)-oxo moiety in compound I (Cpd I), with the ω-1 position being closest, indicating higher reactivity at this site. QM/MM calculations identified hydrogen abstraction as the rate-determining step, with the ω-1S transition state exhibiting the lowest energy barrier, consistent with experimentally observed enantioselectivity. Energy decomposition analysis revealed that variations in quantum mechanical energy (ΔEQM) significantly influence reaction barriers, with the most efficient hydrogen abstraction occurring at the ω-1S and ω-2R positions. These findings underscore the importance of substrate positioning within the active site in determining product selectivity. Comparisons with two related P450s, P450BM3 and P450SPα, further highlighted the critical role of active site architecture and substrate positioning in modulating selectivity. While surrounding residues do not directly dictate product selectivity, they shape the active site environment and influence substrate positioning. Furthermore, our analysis revealed a previously unrecognized catalytic role of Ala299. These findings provide a deeper mechanistic understanding of human CYP2E1 and offer valuable insights for its precise engineering in targeted C-H functionalization.

Initial Conditions for Excited-State Dynamics in Solvated Systems: A Case Study

Simulating excited-state dynamics or computing spectra for molecules in condensed phases requires sampling the ground state to generate initial conditions. Initial conditions (or snapshots for spectra) are typically produced by QM/MM Boltzmann sampling following MM equilibration or optimization. Given the switch from a MM to a QM/MM potential energy surface, one should discard a set period of time (which we call the "healing time") from the beginning of the QM/MM trajectory. Ideally, the healing time is as short as possible (to avoid unnecessary computational effort), but long enough to equilibrate to the QM/MM ground state distribution. Healing times in previous studies range from tens of femtoseconds to tens of picoseconds, suggesting the need for guidelines to choose a healing time. We examine the effect of healing time on the nonadiabatic dynamics and spectrum of a first-generation Donor-Acceptor Stenhouse Adduct in chloroform. Insufficient healing times skew the branching ratio of ground state products and alter the relaxation time for one pathway. The influence of the healing time on the absorption spectrum is less pronounced, warning that the spectrum is not a sensitive indicator for the quality of a set of initial conditions for dynamics. We demonstrate that a reasonable estimate for the healing time can be obtained by monitoring the solute temperature during the healing trajectory. We suggest that this procedure should become standard practice for determining healing times to generate initial conditions for nonadiabatic QM/MM simulations in large molecules and condensed phases.

Mechanistic Perspective on C-N and C-S Bond Construction Catalyzed by Cytochrome P450 Enzymes

Cytochrome P450 enzymes catalyze a large number of oxidative transformations that are responsible for natural product synthesis. Recent studies have revealed their unique ability to catalyze the formation of C-N and C-S bonds, broadening their biosynthetic applications. However, the enzymatic mechanisms behind these reactions are still unclear. This review focuses on theoretical insights into the mechanisms of P450-catalyzed C-N and C-S bond formation. The key roles of the conformational dynamics of substrate radicals, influenced by the enzyme environment, in modulating selectivity and reactivity are highlighted. Understanding these reaction mechanisms offers valuable guidance for P450 enzyme engineering and the design of biosynthetic applications.

Wide transition-state ensemble as key component for enzyme catalysis

Transition-state (TS) theory has provided the theoretical framework to explain the enormous rate accelerations of chemical reactions by enzymes. Given that proteins display large ensembles of conformations, unique TSs would pose a huge entropic bottleneck for enzyme catalysis. To shed light on this question, we studied the nature of the enzymatic TS for the phosphoryl-transfer step in adenylate kinase by quantum-mechanics/molecular-mechanics calculations. We find a structurally wide set of energetically equivalent configurations that lie along the reaction coordinate and hence a broad transition-state ensemble (TSE). A conformationally delocalized ensemble, including asymmetric TSs, is rooted in the macroscopic nature of the enzyme. The computational results are buttressed by enzyme kinetics experiments that confirm the decrease of the entropy of activation predicted from such wide TSE. TSEs as a key for efficient enzyme catalysis further boosts a unifying concept for protein folding and conformational transitions underlying protein function.

Impact of Fluctuations in the Peridinin-Chlorophyll a-Protein on the Energy Transfer: Insights from Classical and QM/MM Molecular Dynamics Simulations

The peridinin-chlorophyll a-protein is a light-harvesting complex found in dinoflagellates, which has an unusually high fraction of carotenoids. The carotenoids are directly involved in the energy transfer to chlorophyll with high efficiency. The detailed mechanism of energy transfer and the roles of the protein in the process remain debated in the literature, in part because most calculations have focused on a limited number of chromophore structures. Here we investigate the magnitude of the fluctuations of the site energies of individual and coupled chromophores, as the results are essential to the understanding of experimental spectra and the energy transfer mechanism. To this end, we sampled conformations of the PCP complex by means of classical and quantum mechanical/molecular mechanical (QM/MM) molecular dynamics simulations. Subsequently we performed (supermolecular) excitation energy calculations on a statistically significant number of snapshots using TD-LC-DFT/CAM-B3LYP and the semiempirical time-dependent long-range corrected density functional tight binding (TD-LC-DFTB2) as the QM method. We observed that the magnitude of the site energy fluctuations is large compared to the differences of the site energies between the chromophores, and this also holds for the coupled chromophores. We also investigated the composition of the coupled states, the effect of coupling on the absorption spectra, as well as transition dipole moment orientations and the possibility of delocalized states with Chl a. Our study thus complements previous computational studies relying on a single structure and establishes the most prominent features of the coupled chromophores that are essential to the robustness of the energy transfer process.

A Robust and Versatile QM/MM Interface for Molecular Dynamics in GROMOS

The integration of quantum mechanics and molecular mechanics (QM/MM) within molecular dynamics simulations is crucial to accurately model complex biochemical systems. Here, we present an enhanced implementation of the QM/MM interface in the GROMOS simulation package, introducing significant improvements in functionality and user control. We present new features, including the link atom scheme, which allows the modeling of QM regions as a part of bigger molecules. Benchmark tests on various systems, including QM water in water, amino acids in water, and tripeptides validate the reliability of the new functionalities. Performance evaluations demonstrate that the updated implementation is efficient, with the primary computational burden attributed to the QM program rather than the QM/MM interface or the MD program itself. The improved QM/MM interface enables more advanced investigations into biomolecular reactivity, enzyme catalysis, and other phenomena requiring detailed quantum mechanical treatment within classical simulations. This work represents a significant advancement in the capabilities of GROMOS, providing enhanced tools to explore complex molecular systems.

Conformational heterogeneity and protonation equilibria shape the photocycle branching in channelrhodopsin-2

Channelrhodopsin-2 is a photoactive membrane protein serving as an ion channel, gathering significant interest for its applications in optogenetics. Despite extensive investigation, several aspects of its photocycle remain elusive and continue to be subjects of ongoing debate. Of particular interest are the localization of the P480 intermediate within the photocycle and the timing of the deprotonation of glutamic acid E90, a critical residue for ChR2 functioning. In this study, we explore the possibility of an early-P480 state, formed directly upon photoillumination of the dark-adapted state, where E90 is deprotonated, as hypothesized in a previous work [Kuhne et al. Proc. Natl. Acad. Sci. 116.19 (2019): 9380]. Employing extended molecular dynamics simulations, deprotonation free energy calculations, and the computation of the infrared band associated with E90, we provide support to the photocycle model proposed by Kuhne et al. Furthermore, our findings show that E90 protonation state is influenced by diverse interconnected variables and provide molecular detail insights that connect E90 interaction pattern with its deprotonation propensity. Our data demonstrate in fact that both protonated and deprotonated E90 are possible in P480 depending on E90 hydrogen bonding pattern and explaining the molecular mechanism at the basis of P480 accumulation under continuous illumination.

How Accurate Are QM/MM Models?

Despite the success and widespread use of QM/MM methods in modeling (bio)chemically important processes, their accuracy is still not well understood. A key reason is because these methods are ultimately approximations to direct QM calculations of very large systems, which are impractical to perform in most cases. We highlight recent progress toward the development of realistic model systems where it is possible to obtain full QM reference data to directly and systematically evaluate the effectiveness of different QM/MM generation schemes. These model systems are highly flexible and can be tailored to probe the sensitivity of a QM/MM model to different reaction types and simulation parameters such as pairing of QM and MM potentials, QM region size, and composition. It is envisaged that this strategy could be used to directly validate different QM/MM generation schemes and spur the development of more robust models in the future.

DNA Triplet Energies by Free Energy Perturbation Theory

Determining the energetics of triplet electronic states of nucleobases in the biological macromolecular environment of nucleic acids is essential for an accurate description of the mechanism of photosensitization and the design of drugs for cancer treatment. In this work, we aim at developing a methodological approach to obtain accurate free energies of triplets in DNA beyond the state of the art, able to reproduce the decrease of triplet energies measured experimentally for T in DNA (270 kJ/mol) vs in the isolated nucleotide in aqueous solution (310 kJ/mol). For such purposes, we adapt the free energy perturbation method to compute the free energy related to the transformation of a pure singlet state into a pure triplet state via "alchemical" intermediates with mixed singlet-triplet nature. By this means, standard deviation errors are only a few kJ/mol, contrary to the large errors of tenths of kJ/mol obtained by averaging the singlet and triplet energies derived from molecular dynamics simulations. The reduced statistical errors obtained by the free energy perturbation approach allow us to rationalize with confidence the triplet stabilization observed experimentally when comparing the thymine nucleotide and thymine in DNA. Spin polarization rather than excimer interactions between the π-stacked nucleobases originates the lower values of the triplet energies in DNA. The developed approach implemented in QM3 shall be useful for determining free energies of triplets and other states like ionic or charge separation states in any other macromolecular system with impact in biomedicine and materials science.

Asymmetric multi-component trifunctionalization reactions with α-Halo Rh-carbenes

Multi-component multi-functionalization reactions involving active intermediates are powerful tools for rapidly generating a wide array of compounds. Metal carbynoids, with their distinct reactivity, hold great promise for developing synthetic methodologies. However, their application in catalytic transfer reactions has been hindered by the limited availability of suitable precursors. In this study, we investigate the catalytic potential of α-halo Rh-carbenes, leveraging the concept of metal carbynoids in multi-functionalization reactions. Through a chiral phosphoric acid-catalyzed asymmetric trifunctionalization, we have developed a method for synthesizing a variety of chiral α-cyclic ketal β-amino esters with high yields and excellent enantioselectivity. Our extensive experimental and computational studies reveal that α-halo Rh-carbenes exhibit carbynoid properties, which facilitate the transformation into functionalized Fischer-type Rh-carbenes through the decomposition of the C-halo bond.

Simulating Ionized States in Realistic Chemical Environments with Algebraic Diagrammatic Construction Theory and Polarizable Embedding

Theoretical simulations of electron detachment processes are vital for understanding chemical redox reactions, semiconductor and electrochemical properties, and high-energy radiation damage. However, accurate calculations of ionized electronic states are very challenging due to their open-shell nature, importance of electron correlation effects, and strong interactions with chemical environment. In this work, we present an efficient approach based on algebraic diagrammatic construction theory with polarizable embedding that allows to accurately simulate ionized electronic states in condensed-phase or biochemical environments (PE-IP-ADC). We showcase the capabilities of PE-IP-ADC by computing the vertical ionization energy (VIE) of thymine molecule solvated in bulk water. Our results show that the second- and third-order PE-IP-ADC methods combined with the basis of set of triple-ζ quality yield a solvent-induced shift in VIE of -0.92 and -0.93 eV, respectively, in an excellent agreement with experimental estimate of -0.9 eV. This work demonstrates the power of PE-IP-ADC approach for simulating charged electronic states in realistic chemical environments and motivates its further development.

Machine Learning Quantum Mechanical/Molecular Mechanical Potentials: Evaluating Transferability in Dihydrofolate Reductase-Catalyzed Reactions

Integrating machine learning potentials (MLPs) with quantum mechanical/molecular mechanical (QM/MM) free energy simulations has emerged as a powerful approach for studying enzymatic catalysis. However, its practical application has been hindered by the time-consuming process of generating the necessary training, validation, and test data for MLP models through QM/MM simulations. Furthermore, the entire process needs to be repeated for each specific enzyme system and reaction. To overcome this bottleneck, it is required that trained MLPs exhibit transferability across different enzyme environments and reacting species, thereby eliminating the need for retraining with each new enzyme variant. In this study, we explore this potential by evaluating the transferability of a pretrained ΔMLP model across different enzyme mutations within the MM environment using the QM/MM-based ML architecture developed by Pan, X. J. Chem. Theory Comput. 2021, 17(9), 5745-5758. The study includes scenarios such as single point substitutions, a homologous enzyme from different species, and even a transition to an aqueous environment, where the last two systems have MM environment that is substantially different from that used in MLP training. The results show that the ΔMLP model effectively captures and predicts the effects of enzyme mutations on electrostatic interactions, producing reliable free energy profiles of enzyme-catalyzed reactions without the need for retraining. The study also identified notable limitations in transferability, particularly when transitioning from enzyme to water-rich MM environments. Overall, this study demonstrates the robustness of the Pan et al.'s QM/MM-based ML architecture for application to diverse enzyme systems, as well as the need for further research and the development of more sophisticated MLP models and training methods.

Mixed atomistic-implicit quantum/classical approach to molecular nanoplasmonics

A multiscale quantum mechanical (QM)/classical approach is presented that is able to model the optical properties of complex nanostructures composed of a molecular system adsorbed on metal nanoparticles. The latter is described by a combined atomistic-continuum model, where the core is described using the implicit boundary element method (BEM) and the surface retains a fully atomistic picture and is treated employing the frequency-dependent fluctuating charge and fluctuating dipole (ωFQFμ) approach. The integrated QM/ωFQFμ-BEM model is numerically compared with state-of-the-art fully atomistic approaches, and the quality of the continuum/core partition is evaluated. The method is then extended to compute surface-enhanced Raman scattering within a time-dependent density functional theory framework.

Accurate QM/MM Molecular Dynamics for Periodic Systems in GPU4PySCF with Applications to Enzyme Catalysis

We present an implementation of the quantum mechanics/molecular mechanics (QM/MM) method for periodic systems using GPU accelerated QM methods, a distributed multipole formulation of the electrostatics, and a pseudobond treatment of the QM/MM boundary. We demonstrate that our method has well-controlled errors, stable self-consistent QM convergence, and energy-conserving dynamics. We further describe an application to the catalytic kinetics of chorismate mutase. Using an accurate hybrid functional reparametrized to coupled cluster energetics, our QM/MM simulations highlight the sensitivity in the calculated rate to the choice of quantum method, quantum region selection, and local protein conformation. Our work is provided through the open-source PySCF package using acceleration from the GPU4PySCF module.

Protein-Driven Electron-Transfer Process in a Fatty Acid Photodecarboxylase

Naturally occurring photoenzymes are rare in nature, but among them, fatty acid photodecarboxylases derived from Chlorella variabilis (CvFAPs) have emerged as promising photobiocatalysts capable of performing the redox-neutral, light-induced decarboxylation of free fatty acids (FAs) into C1-shortened n-alka(e)nes. Using a hybrid QM/MM approach combined with a polarizable embedding scheme, we identify the structural changes of the active site and determine the energetic landscape of the forward electron transfer (fET) from the FA substrate to the excited flavin adenine dinucleotide. We obtain a charge-transfer diradical structure where a water molecule rearranges spontaneously to form a H-bond interaction with the excited flavin, while the FA's carboxylate group twists and migrates away from it. Together, these structural modifications provide the driving force necessary for the fET to proceed in a downhill direction. Moreover, by examining the R451K mutant where the FA substrate is farther from the flavin core, we show that the marked reduction of the electronic coupling is counterbalanced by an increased driving force, resulting in a fET lifetime similar to the WT, thereby suggesting a resilience of the process to this mutation. Finally, through QM/MM molecular dynamic simulations, we reveal that, following fET, the decarboxylation of the FA radical occurs within tens of picoseconds, overcoming an energy barrier of ∼0.1 eV. Overall, by providing an atomistic characterization of the photoactivation of CvFAP, this work can be used for future protein engineering.

Efficient Polarizable QM/MM Using the Direct Reaction Field Hamiltonian with Electrostatic Potential Fitted Multipole Operators

Electronic polarization and dispersion are decisive actors in determining interaction energies between molecules. These interactions have a particularly profound effect on excitation energies of molecules in complex environments, especially when the excitation involves a significant degree of charge reorganization. The direct reaction field (DRF) approach, which has seen a recent revival of interest, provides a powerful framework for describing these interactions in quantum mechanics/molecular mechanics (QM/MM) models of systems, where a small subsystem of interest is described using quantum chemical methods and the remainder is treated with a simple MM force field. In this paper we show how the DRF approach can be combined with the electrostatic potential fitted (ESPF) multipole operator description of the QM region charge density, which significantly improves the efficiency of the method, particularly for large MM systems, and for typical calculations effectively eliminates the dependence on MM system size. We also show how the DRF approach can be combined with fluctuating charge descriptions of the polarizable environment, as well as previously used atom-centered dipole-polarizability based models. We further show that the ESPF-DRF method provides an accurate description of molecular interactions in both ground and excited electronic states of the QM system and apply it to predict the gas to aqueous solution solvatochromic shifts in the UV/visible absorption spectrum of acrolein.

Toward a Generalizable Machine-Learned Potential for Metal-Organic Frameworks

Machine-learned potentials (MLPs) have transformed the field of molecular simulations by scaling "quantum-accurate" potentials to linear time complexity. While they provide more accurate reproduction of physical properties as compared to empirical force fields, it is still computationally costly to generate their training data sets from ab initio calculations. Despite the emergence of foundational or general MLPs for organic molecules and dense materials, it is unexplored if one general MLP can be effectively developed for a wide variety of nanoporous metal-organic frameworks (MOFs) with different chemical moieties and geometric properties. Herein, by leveraging upon data-efficient equivariant MLPs, we demonstrate the possibility of developing a general MLP for nearly 3000 Zn-based MOFs. After curating a training data set comprising augmented MOF structures generated from density functional theory optimization, we validate the reliability of the general MLP in predicting accurate forces and energies when evaluated on a test set with chemically distinct MOF structures. Despite incurring slightly higher errors on structures containing rare chemical moieties, the general MLP can reliably reproduce physical (e.g., vibrational, thermodynamic, and mechanical) properties for a large sample of Zn-based MOFs. Crucially, by developing one MLP for many MOFs, the computational cost of high-throughput screening is potentially reduced by a few orders of magnitude. This enables us to predict quantum-accurate properties for notable Zn-MOFs that were previously uninvestigated via expensive theoretical calculations. To facilitate computational discovery among other families of complex chemical structures, we provide our data set and codes in the public Zenodo repository.

Deep Learning-Driven Insights into Enzyme-Substrate Interaction Discovery

Enzymes are ubiquitous catalysts with enormous application potential in biomedicine, green chemistry, and biotechnology. However, accurately predicting whether a molecule serves as a substrate for a specific enzyme, especially for novel entities, remains a significant challenge. Compared with traditional experimental methods, computational approaches are much more resource-efficient and time-saving, but they often compromise on accuracy. To address this, we introduce the molecule-enzyme interaction (MEI) model, a novel machine learning framework designed to predict the probability that a given molecule is a substrate for a specified enzyme with high accuracy. Utilizing a comprehensive data set that encapsulates extensive information on enzymatic reactions and enzyme sequences, the MEI model seamlessly combines atomic environmental data with amino acid sequence features through an advanced attention mechanism within a hierarchical neural network. Empirical evaluations have confirmed that the MEI model outperforms the current state-of-the-art model by at least 6.7% in prediction accuracy and 8.5% in AUROC, underscoring its enhanced predictive capabilities. Additionally, the MEI model demonstrates remarkable generalization across data sets of varying qualities and sizes. This adaptability is further evidenced by its successful application in diverse areas, such as predicting interactions within the CYP450 enzyme family and achieving an outstanding accuracy of 90.5% in predicting the enzymatic breakdown of complex plastics within environmental applications. These examples illustrate the model's ability to effectively transfer knowledge from coarsely annotated enzyme databases to smaller, high-precision data sets, robustly modeling both sparse and high-quality databases. We believe that this versatility firmly establishes the MEI model as a foundational tool in enzyme research with immense potential to extend beyond its original scope.

A Polarizable CASSCF/MM Approach Using the Interface Between OpenMMPol Library and Cfour

We present a polarizable embedding quantum mechanics/molecular mechanics (QM/MM) framework for ground- and excited-state Complete Active Space Self-Consistent Field (CASSCF) calculations on molecules within complex environments, such as biological systems. These environments are modeled using the AMOEBA polarizable force field. This approach is implemented by integrating the OpenMMPol library with the CFour quantum chemistry software suite. The implementation supports both single-point energy evaluations and geometry optimizations, facilitated by the availability of analytical gradients. We demonstrate the methodology by applying it to two distinct photoreceptors, exploring the impact of the protein environment on the structural and photophysical properties of their embedded chromophores.

QM/MM Calculations on Excited-State Proton Transfer and Photoisomerization of a Red Fluorescent Protein mKeima with Large Stokes Shift

Large Stokes shift red fluorescent proteins (LSS-RFPs) are of growing interest for multicolor bioimaging applications. However, their photochemical mechanisms are not fully understood. Here, we employed the QM(XDW-CASPT2//CASSCF)/MM method to investigate the excited-state proton transfer and photoisomerization processes of the LSS-RFP mKeima starting from its cis neutral isomer. Upon excitation to the bright S1 state in the Franck-Condon region, mKeima relaxes to a metastable minimum-energy state. From this short-lived species, two competing deactivation pathways are available: the excited-state proton transfer in the S1 state, and the S1 decay via the S1/S0 conical intersection as a result of the cis-trans photoisomerization. In comparison, the former is a dominant excited-state relaxation pathway, leading to the cis anionic isomer of mKeima in the S1 state. This anionic intermediate then undergoes cis-trans photoisomerization after overcoming a barrier of approximately 10 kcal/mol in the S1 state, which is followed by an excited-state decay via the S1/S0 conical intersection region. The efficient nonadiabatic decay of the cis anionic isomer of mKeima in the S1 state inhibits the radiative process, leading to a weak emission around 520 nm observed experimentally. These findings shed important mechanistic light on the experimental observations and provide valuable insights that could help in the design of LSS-RFPs with superior fluorescence properties.

Catalytic Role of Methanol in Anodic Coupling Reactions Involving Alcohol Trapping of Cation Radicals

In anodic electrosynthesis, cation radicals are often key intermediates that can be highly susceptible to nucleophilic attack and/or deprotonation, with the selectivity of competing pathways dictating product yield. In this work, we computationally investigate the role of methanol in alcohol trapping of enol ether cation radicals for which substantial modulation of the reaction yield by the solvent environment was previously observed. Reaction free energies computed for intramolecular coupling unequivocally demonstrate that the key intramolecular alcohol attack on the oxidized enol ether group is catalyzed by methanol, proceeding through overall second-order kinetics. Methanol complexation with the formed oxonium ion group gives rise to a "Zundel-like", shared proton conformation, providing a critical driving force for the intramolecular alcohol attack. Free energies computed for methanol solvent attack of enol ether cation radicals demonstrate an analogous mechanism and overall third-order kinetics, due to similar complexation from a secondary methanol molecule to form the "Zundel-like", shared proton conformation. As catalyzed by methanol, both intramolecular alcohol attack and methanol attack on the oxidized enol ether group are barrierless or low-barrier reactions, with kinetic competition dictated by the conformational free energy profile of the cation radical substrate and the difference in reaction rate orders.

Investigating the Origin of Automatic Rhodopsin Modeling Outliers Using the Microbial Gloeobacter Rhodopsin as Testbed

The automatic rhodopsin modeling (ARM) approach is a computational workflow devised for the automatic buildup of hybrid quantum mechanics/molecular mechanics (QM/MM) models of wild-type rhodopsins and mutants, with the purpose of establishing trends in their photophysical and photochemical properties. Despite the success of ARM in accurately describing the visible light absorption maxima of many rhodopsins, for a few cases, called outliers, it might lead to large deviations with respect to experiments. Applying ARM toGloeobacter rhodopsin (GR), a microbial rhodopsin with important applications in optogenetics, we analyze the origin of such outliers in the absorption energies obtained for GR wild-type and mutants at neutral pH, with a total root-mean-square deviation (RMSD) of 0.42 eV with respect to the experimental GR excitation energies. Having discussed the importance and the uncertainty of one particular amino-acid pKa, namely histidine at position 87, we propose and test several modifications to the standard ARM protocol: (i) improved pKa predictions along with the consideration of several protonation microstates, (ii) attenuation of the opsin electrostatic potential at short-range, (iii) substitution of the state-average complete active space (CAS) electronic structure method by its state-specific approach, and (iv) complete replacement of CAS with mixed-reference spin-flip time-dependent density functional theory (MRSF-TDDFT). The best RMSD result we obtain is 0.2 eV combining the protonation of H87 and using MRSF/CAMH-B3LYP.

Local Wave Function Embedding: Correlation Regions in PNO-LCCSD(T)-F12 Calculations

Many chemical reactions affect only a rather small number of bonds, leaving the largest part of the chemical and geometrical structure of the molecules nearly unchanged. In this work we extended the previously proposed region method [J. Chem. Phys. 128, 144106 (2008)] to PNO-LCCSD(T)-F12. Using this method, we investigate whether accurate reaction energies for larger systems can be obtained by correlating only the electrons in a region of localized molecular orbitals close to the reaction center at high-level (PNO-LCCSD(T)-F12). The remainder is either treated at lower level (PNO-LMP2-F12) or left uncorrelated (Hartree-Fock frozen core). It is demonstrated that indeed the computed reaction energies converge rather quickly with the size of the correlation regions toward the results of the full calculations. Typically, 2-3 bonds from the reacting atoms need to be included to reproduce the results of the full calculations to within ±0.2 kcal/mol. We also computed spin-state energy differences in a large transition metal complex, where a factor of 15 in computation time could be saved, still yielding a result that is within ±0.1 kcal/mol of the one obtained in a full PNO-LCCSD(T)-F12 calculation.

Emerging trends in plant natural products biosynthesis: a chemical perspective

Biosynthetic pathways are multistep processes transforming simple substrates into more complex structures. Over the past two decades, our understanding of these pathways, especially for specialized plant metabolites, has significantly increased. This surge is due to numerous scientific advancements such as next-generation sequencing, improved analytical platforms, and metabolite-transcript networks. The uprising of data sharing through public databases has also fostered collaboration and knowledge dissemination. Growing concerns about the supply of therapeutic natural products and their environmental impact have led to exploring sustainable alternatives like heterologous expression, which requires extensive knowledge of these pathways. Herein, we review emerging approaches in biosynthetic pathway elucidations and their prospects for their efficient integration.

Multiscale calculations reveal new insights into the reaction mechanism between KRASG12C and α, β-unsaturated carbonyl of covalent inhibitors

Utilizing α,β-unsaturated carbonyl group as Michael acceptors to react with thiols represents a successful strategy for developing KRASG12C inhibitors. Despite this, the precise reaction mechanism between KRASG12C and covalent inhibitors remains a subject of debate, primarily due to the absence of an appropriate residue capable of deprotonating the cysteine thiol as a base. To uncover this reaction mechanism, we first discussed the chemical reaction mechanism in solvent conditions via density functional theory (DFT) calculation. Based on this, we then proposed and validated the enzymatic reaction mechanism by employing quantum mechanics/molecular mechanics (QM/MM) calculation. Our QM/MM analysis suggests that, in biological conditions, proton transfer and nucleophilic addition may proceed through a concerted process to form an enolate intermediate, bypassing the need for a base catalyst. This proposed mechanism differs from previous findings. Following the formation of the enolate intermediate, solvent-assisted tautomerization results in the final product. Our calculations indicate that solvent-assisted tautomerization is the rate-limiting step in the catalytic cycle under biological conditions. On the basis of this reaction mechanism, the calculated kinact/ki for two inhibitors is consistent well with the experimental results. Our findings provide new insights into the reaction mechanism between the cysteine of KRASG12C and the covalent inhibitors and may provide valuable information for designing effective covalent inhibitors targeting KRASG12C and other similar targets.

Advances and Challenges in Milestoning Simulations for Drug-Target Kinetics

Molecular dynamics simulations have become indispensable for exploring complex biological processes, yet their limitations in capturing rare events hinder our understanding of drug-target kinetics. In this Perspective, we investigate the domain of milestoning simulations to understand this challenge. The milestoning approach divides the phase space of the drug-target complex into discrete cells, offering extended time scale insights. This Perspective traces the history, applications, and future potential of milestoning simulations in the context of drug-target kinetics. It explores the fundamental principles of milestoning, highlighting the importance of probabilistic transitions and transition time independence. Markovian milestoning with Voronoi tessellations is revisited to address the traditional milestoning challenges. While observing the advancements in this field, this Perspective also addresses impending challenges in estimating drug-target unbinding rate constants through milestoning simulations, paving the way for more effective drug design strategies.