Enhanced Sampling Techniques Research Articles

Revisiting the Enhanced Chemical Reactivity in Water Microdroplets: The Case of a Diels-Alder Reaction.

Often, chemical reactions are markedly accelerated in microdroplets compared with the corresponding bulk phase. While identifying the precise causative factors remains challenging, the interfacial electric field (IEF) and partial solvation are the two widely proposed factors, accounting for the acceleration or turning on of many reactions in microdroplets. In sharp contrast, this combined computational and experimental study demonstrates that these two critical factors have a negligible effect on promoting a model Diels-Alder (DA) reaction between cyclopentadiene and acrylonitrile in water microdroplets. Instead, the acceleration of the DA reaction appears to be driven by the effect of confinement and the concentration increase caused by evaporation. Quantum chemical calculations and ab initio molecular dynamics simulations coupled with enhanced sampling techniques predict that the air-water interface exhibits a higher free-energy barrier of this reaction than the bulk, while external electric fields marginally reduce the barrier. Remarkably, the catalytic capability of the IEF at the water microdroplet surface is largely hampered by its fluctuating character. Mass spectrometric assessment of the microdroplet reaction corroborates these findings, suggesting that the DA reaction is not facilitated by the IEF as increasing the spray potential suppresses the DA products by promoting substrate oxidation. While the DA reaction exhibits a surface preference in water microdroplets, the same reaction tends to occur mainly within the core of the acetonitrile microdroplet, suggesting that the partial solvation is not necessarily a critical factor for accelerating this reaction in microdroplets. Moreover, experiments indicate that the rapid evaporation of microdroplets and subsequent reagent enrichment within the accessible confined volume of microdroplets caused the observed acceleration of the DA reaction in water microdroplets.

Investigating Ligand-Mediated Conformational Dynamics of Pre-miR21: A Machine-Learning-Aided Enhanced Sampling Study.

MicroRNAs (miRs) are short, noncoding RNA strands that regulate the activity of mRNAs by affecting the repression of protein translation, and their dysregulation has been implicated in several pathologies. miR21 in particular has been implicated in tumorigenesis and anticancer drug resistance, making it a critical target for drug design. miR21 biogenesis involves precise biochemical pathways, including the cleavage of its precursor, pre-miR21, by the enzyme Dicer. The present work investigates the conformational dynamics of pre-miR21, focusing on the role of adenine29 in switching between Dicer-binding-prone and inactive states. We also investigated the effect of L50, a cyclic peptide binder of pre-miR21 and a weak inhibitor of its processing. Using time series data and our novel collective variable-based enhanced sampling technique, OneOPES, we simulated these conformational changes and assessed the effect of L50 on the conformational plasticity of pre-miR21. Our results provide insight into peptide-induced conformational changes and pave the way for the development of a computational platform for the screening of inhibitors of pre-miR21 processing that considers RNA flexibility, a stepping stone for effective structure-based drug design, with potentially broad applications in drug discovery.

Selective Protonation of Catalytic Dyad for γ-Secretase-Mediated Hydrolysis Revealed by Multiscale Simulations.

γ-Secretase plays a crucial role in producing disease-related amyloid-β proteins by cleaving the amyloid precursor protein (APP). The enzyme employs its catalytic dyad containing two aspartates (Asp257 and Asp385) to hydrolyze the substrate by a general acid-base catalytic mechanism, necessitating monoprotonation of the two aspartates for efficient hydrolysis. However, the precise protonation states of the aspartates remain uncertain. In this study, we employed a multiscale computational approach to investigate the dependence of the catalytic efficiency of γ-secretase on the protonation states of its catalytic dyad. Over 200 ms unbiased atomistic simulations of the substrate-enzyme complex reveal diverse orientations of the scissile bond of the bound substrate and accessible structural ensembles of the catalytic dyad with Asp257-Asp385 distances fluctuating between 4 and 10 Å. With a quantum mechanics/molecular mechanics (QM/MM) approach accelerated by enhanced sampling techniques, we find that the first step of the hydrolysis reaction, i.e., the formation of a gem-diol intermediate, experiences a higher reaction barrier by ∼2 kcal/mol when Asp385 is protonated. Furthermore, we find that Arg269 of the enzyme is most likely responsible for this preference of the protonation state: its basic side chain is spatially close to that of Asp257 and specifically stabilizes the transition state electrostatically when Asp257 is protonated. Collectively, our study suggests that Asp257 is likely the favored protonation site for APP cleavage by γ-secretase.

Adaptive CVgen: Leveraging reinforcement learning for advanced sampling in protein folding and chemical reactions

Enhanced sampling techniques have traditionally encountered two significant challenges: identifying suitable reaction coordinates and addressing the exploration-exploitation dilemma, particularly the difficulty of escaping local energy minima. Here, we introduce Adaptive CVgen, a universal adaptive sampling framework designed to tackle these issues. Our approach utilizes a set of collective variables (CVs) to comprehensively cover the system's potential evolutionary phase space, generating diverse reaction coordinates to address the first challenge. Moreover, we integrate reinforcement learning strategies to dynamically adjust the generated reaction coordinates, thereby effectively balancing the exploration-exploitation dilemma. We apply this framework to sample the conformational space of six proteins transitioning from completely disordered states to folded states, as well as to model the chemical synthesis process of C60, achieving conformations that perfectly match the standard C60 structure. The results demonstrate Adaptive CVgen's effectiveness in exploring new conformations and escaping local minima, achieving both sampling efficiency and exploration accuracy. This framework holds potential for extending to various related challenges, including protein folding dynamics, drug targeting, and complex chemical reactions, thereby opening promising avenues for application in these fields.

Quantum mechanics/molecular mechanics (QM/MM) methods in drug design: a comprehensive review of development and applications

The integration of quantum mechanics (QM) and molecular mechanics (MM) methods, known as QM/MM, has emerged as a powerful computational approach in drug design. This hybrid technique combines the accuracy of QM calculations for the reactive region with the computational efficiency of MM methods for the surrounding environment. QM/MM methods have proven invaluable in studying chemical reactions, exploring enzyme mechanisms, and investigating ligand-protein interactions, all of which are crucial for rational drug design. This review provides a comprehensive overview of the QM/MM methodology, its theoretical foundations, and its applications in various aspects of drug design. We discussed the key components, such as QM and MM region partitioning, link atom schemes, and boundary treatments. Additionally, we highlight recent advancements and challenges in QM/MM methods, including polarizable force fields, implicit solvation models, and enhanced sampling techniques. Furthermore, we present illustrative examples showcasing the successful application of QM/MM methods in lead optimization, virtual screening, and the elucidation of biochemical mechanisms relevant to drug targets. Finally, we provide perspectives on future developments and the potential impact of QM/MM methods on the drug discovery pipeline.

Elucidating Protein Structures in the Gas Phase: Traversing Configuration Space with Biasing Methods.

Achieving accurate characterization of protein structures in the gas phase continues to be a formidable challenge. To tackle this issue, the present study employs Molecular Dynamics (MD) simulations in tandem with enhanced sampling techniques (methods designed to efficiently explore protein conformations). The objective is to identify suitable structures of proteins by contrasting their calculated Collision Cross-Section (CCS) with those observed experimentally. Significant discrepancies were observed between the initial MD-simulated and experimentally measured CCS values through Ion Mobility-Mass Spectrometry (IMS-MS). To bridge this gap, we employed two distinct enhanced sampling methods, Harmonic Biasing Potential and Adaptive Biasing Force, which help the proteins overcome energy barriers to adopt more compact configurations. These techniques leverage the radius of gyration as a reaction coordinate (guiding parameter), guiding the system toward compressed states that potentially match experimental configurations more closely. The guiding forces are only employed to overcome existing barriers and are removed to allow the protein to naturally arrive at a potential gas phase configuration. The results demonstrated close alignment (within ∼4%) between simulated and experimental CCS values despite using different strengths and/or methods, validating their efficacy. This work lays the groundwork for future studies aimed at optimizing biasing methods and expanding the collective variables used for more accurate gas-phase structural predictions.

A direct computational assessment of vinculin-actin unbinding kinetics reveals catch bonding behavior.

Vinculin forms a catch bond with the cytoskeletal polymer actin, displaying an increased bond lifetime upon force application. Notably, this behavior depends on the direction of the applied force, which has significant implications for cellular mechanotransduction. In this study, we present a comprehensive molecular dynamics simulation study, employing enhanced sampling techniques to investigate the thermodynamic, kinetic, and mechanistic aspects of this phenomenon at physiologically relevant forces. We dissect a catch bond mechanism in which force shifts vinculin between either a weakly- or strongly-bound state. Our results demonstrate that models for these states have unbinding times consistent with those from single-molecule studies, and suggest that both have some intrinsic catch bonding behavior. We provide atomistic insight into this behavior, and show how a directional pulling force can promote the strong or weak state. Crucially, our strategy can be extended to capture the difficult-to-capture effects of small mechanical forces on biomolecular systems in general, and those involved in mechanotransduction more specifically.

Enhanced Sampling of Biomolecular Slow Conformational Transitions Using Adaptive Sampling and Machine Learning.

Biomolecular simulations often suffer from the "time scale problem", hindering the study of rare events occurring over extended time scales. Enhanced sampling techniques aim to alleviate this issue by accelerating conformational transitions, yet they typically necessitate well-defined collective variables (CVs), posing a significant challenge. Machine learning offers promising solutions but typically requires rich training data encompassing the entire free energy surface (FES). In this work, we introduce an automated iterative pipeline designed to mitigate these limitations. Our protocol first utilizes a CV-free count-based adaptive sampling method to generate a data set rich in rare events. From this data set, slow modes are identified using Koopman-reweighted time-lagged independent component analysis (KTICA), which are subsequently leveraged by on-the-fly probability enhanced sampling (OPES) to efficiently explore the FES. The effectiveness of our pipeline is demonstrated and further compared with the common Markov State Model (MSM) approach on two model systems with increasing complexity: alanine dipeptide (Ala2) and deca-alanine (Ala10), underscoring its applicability across diverse biomolecular simulations.

Interfacial Ice Density Fluctuations Inform Surface Ice-Philicity.

The propensity of a surface to nucleate ice or bind to ice is governed by its ice-philicity─its relative preference for ice over liquid water. However, the relationship between the features of a surface and its ice-philicity is not well understood, and for surfaces with chemical or topographical heterogeneity, such as proteins, their ice-philicity is not even well-defined. In the analogous problem of surface hydrophobicity, it has been shown that hydrophobic surfaces display enhanced low water-density (vapor-like) fluctuations in their vicinity. To interrogate whether enhanced ice-like fluctuations are similarly observed near ice-philic surfaces, here we use molecular simulations and enhanced sampling techniques. Using a family of model surfaces for which the wetting coefficient, k, has previously been characterized, we show that the free energy of observing rare interfacial ice-density fluctuations decreases monotonically with increasing k. By utilizing this connection, we investigate a set of fcc systems and find that the (110) surface is more ice-philic than the (111) or (100) surfaces. By additionally analyzing the structure of interfacial ice, we find that all surfaces prefer to bind to the basal plane of ice, and the topographical complementarity of the (110) surface to the basal plane explains its higher ice-philicity. Using enhanced interfacial ice-like fluctuations as a measure of surface ice-philicity, we then characterize the ice-philicity of chemically heterogeneous and topologically complex systems. In particular, we study the spruce budworm antifreeze protein (sbwAFP), which binds to ice using a known ice-binding site (IBS) and resists engulfment using nonbinding sites of the protein (NBSs). We find that the IBS displays enhanced interfacial ice-density fluctuations and is therefore more ice-philic than the two NBSs studied. We also find the two NBSs are similarly ice-phobic. By establishing a connection between interfacial ice-like fluctuations and surface ice-philicity, our findings thus provide a way to characterize the ice-philicity of heterogeneous surfaces.

Efficient machine learning approach for accurate free-energy profiles and kinetic rates.

The computational exploration of reactive processes is challenging due to the requirement of thorough sampling across the free energy landscape using accurate ab initio methods. To address these constraints, machine learning potentials are employed, yet their training for this kind of problem is still a laborious and tedious task. In this study, we present an efficient approach to train these potentials by cleverly using a single batch of unbiased trajectories that avoid the pitfalls of trajectories artificially biased along a suboptimal collective variable. This strategy, when integrated with current enhanced sampling techniques, allows to obtain free energy profiles and kinetic rates of ab initio quality, yet dramatically reducing the computational cost.

Normalizing flows as an enhanced sampling method for atomistic supercooled liquids

Abstract Normalizing flows can transform a simple prior probability distribution into a more complex target distribution. Here, we evaluate the ability and efficiency of generative machine learning methods to sample the Boltzmann distribution of an atomistic model for glass-forming liquids. This is a notoriously difficult task, as it amounts to ergodically exploring the complex free energy landscape of a disordered and frustrated many-body system. We optimize a normalizing flow model to successfully transform high-temperature configurations of a dense liquid into low-temperature ones, near the glass transition. We perform a detailed comparative analysis with established enhanced sampling techniques developed in the physics literature to assess and rank the performance of normalizing flows against state-of-the-art algorithms. We demonstrate that machine learning methods are very promising, showing a large speedup over conventional molecular dynamics. Normalizing flows show performances comparable to parallel tempering and population annealing, while still falling far behind the swap Monte Carlo algorithm. Our study highlights the potential of generative machine learning models in scientific computing for complex systems, but also points to some of its current limitations and the need for further improvement.

A New Route to the Prebiotic Synthesis of Glycine via Ab Initio-Based Machine Learning Calculations.

In this work, we study the synthesis of glycine, the simplest amino acid, using ab initio molecular dynamics and enhanced sampling techniques to explore and quantify novel potential pathways. Our protocol integrates state-of-the-art machine learning approaches, allowing us to sample relevant chemical spaces more efficiently. We discover a novel "oxyglycolate path", distinct from the "standard" Strecker mechanism, identify new intermediates, and provide a full thermodynamic characterization of all reaction steps. This alternative pathway aligns better with meteoritic and experimental observations, paving the way for further investigations. Integrating quantum accuracy and machine learning in prebiotic chemistry represents a methodological milestone advancing the exploration of life's prebiotic origins.

Improving Machine Learned Force Fields for Complex Fluids through Enhanced Sampling: A Liquid Crystal Case Study.

Machine learned force fields offer the potential for faster execution times while retaining the accuracy of traditional DFT calculations, making them promising candidates for molecular simulations in cases where reliable classical force fields are not available. Some of the challenges associated with machine learned force fields include simulation stability over extended periods of time and ensuring that the statistical and dynamical properties of the underlying simulated systems are correctly captured. In this work, we propose a systematic training pipeline for such force fields that leads to improved model quality, compared to that achieved by traditional data generation and training approaches. That pipeline relies on the use of enhanced sampling techniques, and it is demonstrated here in the context of a liquid crystal, which exemplifies many of the challenges that are encountered in fluids and materials with complex free energy landscapes. Our results indicate that, whereas the majority of traditional machine learned force field training approaches lead to molecular dynamics simulations that are only stable over hundred-picosecond trajectories, our approach allows for stable simulations over tens of nanoseconds for organic molecular systems comprising thousands of atoms.

Learning Collective Variables with Synthetic Data Augmentation through Physics-Inspired Geodesic Interpolation.

In molecular dynamics simulations, rare events, such as protein folding, are typically studied using enhanced sampling techniques, most of which are based on the definition of a collective variable (CV) along which acceleration occurs. Obtaining an expressive CV is crucial, but often hindered by the lack of information about the particular event, e.g., the transition from unfolded to folded conformation. We propose a simulation-free data augmentation strategy using physics-inspired metrics to generate geodesic interpolations resembling protein folding transitions, thereby improving sampling efficiency without true transition state samples. This new data can be used to improve the accuracy of classifier-based methods. Alternatively, a regression-based learning scheme for CV models can be adopted by leveraging the interpolation progress parameter.

Registry alteration in Dynein's microtubule-binding domain: A AAA domain-guided event

Dynein, a motor protein, harnesses chemical energy from ATP hydrolysis to generate mechanical output as it travels along microtubular tracks. Essential to this process is the microtubule-binding domain (MTBD), which facilitates the interaction with and detachment from microtubules. Previous studies have proposed that the mechanism governing this interaction is primarily driven by the coiled-coil stalk attached to the MTBD. However, conflicting arguments suggest the presence of two-way communications, where the binding and unbinding mechanisms may also influence the nucleotide state of dynein. In this study, we employed all-atom explicit solvent simulations, enhanced sampling techniques, and coarse-grained methodologies to systematically investigate the effects of stalk and MT on the structural stabilities of different conformations of MTBD and their sequence of events during the transition from one to another. We found that the globular MTBD domain without stalk and MT predominantly resides in its weak binding configuration. Upon introduction of a limited length of stalk and interaction with MT, a balance between the strong and weak binding configurations is restored. Further introduction of the full-length stalk and interaction with MT strengthen the kinetic stability of these two configurations. We have also explored the sequence of events of the relevant transition between these two states using coarse-grained simulation protocols both in the presence and absence of interaction with MT. In the absence of MT interactions, we found a mixed conformational state of the system that corresponds to the already published crystal structure of Dynein. In the presence of MT, the conformational change of the stalk precedes the conformational change of the MTBD globular domain. Our findings support the view that the nucleotide state of the AAA+ ring determines the state of the MTBD domain through stalk, shedding light on the intricate mechanisms governing dynein's interaction with microtubules.

Molecular Bases and Specificity behind the Activation of the Immune System OAS/RNAse L Pathway by Viral RNA.

The first line of defense against invading pathogens usually relies on innate immune systems. In this context, the recognition of exogenous RNA structures is primordial to fight, notably, against RNA viruses. One of the most efficient immune response pathways is based on the sensing of RNA double helical motifs by the oligoadenylate synthase (OAS) proteins, which in turn triggers the activity of RNase L and, thus, cleaves cellular and viral RNA. In this contribution, by using long-range molecular dynamics simulations, complemented with enhanced sampling techniques, we elucidate the structural features leading to the activation of OAS by interaction with a model double-strand RNA oligomer mimicking a viral RNA. We characterize the allosteric regulation induced by the nucleic acid leading to the population of the active form of the protein. Furthermore, we also identify the free energy profile connected to the active vs. inactive conformational transitions in the presence and absence of RNA. Finally, the role of two RNA mutations, identified as able to downregulate OAS activation, in shaping the protein/nucleic acid interface and the conformational landscape of OAS is also analyzed.

Analogies and Differences in the Photoactivation Mechanism of Bathy and Canonical Bacteriophytochromes Revealed by Multiscale Modeling.

Bacteriophytochromes are light-sensing biological machines that switch between two photoreversible states, Pr and Pfr. Their relative stability is opposite in canonical and bathy bacteriophytochromes, but in both cases the switch between them is triggered by the photoisomerization of an embedded bilin chromophore. We applied an integrated multiscale strategy of excited-state QM/MM nonadiabatic dynamics and (QM/)MM molecular dynamics simulations with enhanced sampling techniques to the Agrobacterium fabrum bathy phytochrome and compared the results with those obtained for the canonical phytochrome Deinococcus radiodurans. Contrary to what recently suggested, we found that photoactivation in both phytochromes is triggered by the same hula-twist motion of the bilin chromophore. However, only in the bathy phytochrome, the bilin reaches the final rotated structure already in the first intermediate. This allows a reorientation of the binding pocket in a microsecond time scale, which can propagate through the entire protein causing the spine to tilt.

Application of Generative Artificial Intelligence in Predicting Membrane Partitioning of Drugs: Combining Denoising Diffusion Probabilistic Models and MD Simulations Reduces the Computational Cost to One-Third.

The optimal interaction of drugs with plasma membranes and membranes of subcellular organelles is a prerequisite for desirable pharmacology. Importantly, for drugs targeting the transmembrane lipid-facing sites of integral membrane proteins, the relative affinity of a drug to the bilayer lipids compared to the surrounding aqueous phase affects the partitioning, access, and binding of the drug to the target site. Molecular dynamics (MD) simulations, including enhanced sampling techniques such as steered MD, umbrella sampling (US), and metadynamics, offer valuable insights into the interactions of drugs with the membrane lipids and water in atomistic detail. However, these methods are computationally prohibitive for the high-throughput screening of drug candidates. This study shows that applying denoising diffusion probabilistic models (DDPMs), a generative AI method, to US simulation data reduces the computational cost significantly. Specifically, the models used only partial (one-third) data from the US simulations and reproduced the complete potential of mean force (PMF) profiles for three FDA-approved drugs (β2-adrenergic agonists) and ∼20 biologically relevant chemicals with known experimentally characterized bilayer locations. Intriguingly, the model can predict the solvation-free energies for partitioning and crossing the bilayer, preferred bilayer locations (low-energy well), and orientations of the ligands with high accuracy. The results indicate that DDPMs can be used to characterize the complete membrane partitioning profile of drug molecules using fewer umbrella sampling simulations at select positions along the bilayer normal (z-axis), irrespective of their amphiphilic-lipophilic-cephalophilic characteristics.

Electronic Polarization Leads to a Drier Dewetted State for Hydrophobic Gating in the Big Potassium Channel.

In the hydrophobic gating mechanism proposed for some ion channels, ion permeation is not blocked by the physical dimension of the channel pore but by its dewetted state which constitutes the energetic bottleneck. A major source of uncertainty in the mechanism is that the dewetted state was not observed in experiments and only probed in simulations using nonpolarizable force fields, which do not accurately represent the properties of confined water. Here we analyze hydration of the central cavity in the pore-gate domain of the Big Potassium channel using molecular dynamics and grand canonical Monte Carlo simulations with enhanced sampling techniques. Including polarization leads to a much drier dewetted state and a higher barrier for the transition to the wet state, suggesting more effective hydrophobic gating. The simulations also identify two backbone carbonyls at the bottom of the selectivity filter as good candidates for characterizing the dewetted state using infrared spectroscopies.

Identifying small-molecules binding sites in RNA conformational ensembles with SHAMAN

The rational targeting of RNA with small molecules is hampered by our still limited understanding of RNA structural and dynamic properties. Most in silico tools for binding site identification rely on static structures and therefore cannot face the challenges posed by the dynamic nature of RNA molecules. Here, we present SHAMAN, a computational technique to identify potential small-molecule binding sites in RNA structural ensembles. SHAMAN enables exploring the conformational landscape of RNA with atomistic molecular dynamics simulations and at the same time identifying RNA pockets in an efficient way with the aid of probes and enhanced-sampling techniques. In our benchmark composed of large, structured riboswitches as well as small, flexible viral RNAs, SHAMAN successfully identifies all the experimentally resolved pockets and ranks them among the most favorite probe hotspots. Overall, SHAMAN sets a solid foundation for future drug design efforts targeting RNA with small molecules, effectively addressing the long-standing challenges in the field.