Extended Phase-Space Methods for Enhanced Sampling in Molecular Simulations: A Review

Accelerating Elastic Property Prediction in Fe-C Alloys through Coupling of Molecular Dynamics and Machine Learning

The scarcity of high-quality data presents a major challenge to the prediction of material properties using machine learning (ML) models. Obtaining material property data from experiments is economically cost-prohibitive, if not impossible. In this work, we address this challenge by generating an extensive material property dataset comprising thousands of data points pertaining to the elastic properties of Fe-C alloys. The data were generated using molecular dynamic (MD) calculations utilizing reference-free Modified embedded atom method (RF-MEAM) interatomic potential. This potential was developed by fitting atomic structure-dependent energies, forces, and stress tensors evaluated at ground state and finite temperatures using ab-initio. Various ML algorithms were subsequently trained and deployed to predict elastic properties. In addition to individual algorithms, super learner (SL), an ensemble ML technique, was incorporated to refine predictions further. The input parameters comprised the alloy's composition, crystal structure, interstitial sites, lattice parameters, and temperature. The target properties were the bulk modulus and shear modulus. Two distinct prediction approaches were undertaken: employing individual models for each property prediction and simultaneously predicting both properties using a single integrated model, enabling a comparative analysis. The efficiency of these models was assessed through rigorous evaluation using a range of accuracy metrics. This work showcases the synergistic power of MD simulations and ML techniques for accelerating the prediction of elastic properties in alloys.

Time-dependent principal component analysis: A unified approach to high-dimensional data reduction using adiabatic dynamics

Systematic reduction of the dimensionality is highly demanded in making a comprehensive interpretation of experimental and simulation data. Principal component analysis (PCA) is a widely used technique for reducing the dimensionality of molecular dynamics (MD) trajectories, which assists our understanding of MD simulation data. Here, we propose an approach that incorporates time dependence in the PCA algorithm. In the standard PCA, the eigenvectors obtained by diagonalizing the covariance matrix are time independent. In contrast, they are functions of time in our new approach, and their time evolution is implemented in the framework of Car-Parrinello or Born-Oppenheimer type adiabatic dynamics. Thanks to the time dependence, each of the step-by-step structural changes or intermittent collective fluctuations is clearly identified, which are often keys to provoking a drastic structural transformation but are easily masked in the standard PCA. The time dependence also allows for reoptimization of the principal components (PCs) according to the structural development, which can be exploited for enhanced sampling in MD simulations. The present approach is applied to phase transitions of a water model and conformational changes of a coarse-grained protein model. In the former, collective dynamics associated with the dihedral-motion in the tetrahedral network structure is found to play a key role in crystallization. In the latter, various conformations of the protein model were successfully sampled by enhancing structural fluctuation along the periodically optimized PC. Both applications clearly demonstrate the virtue of the new approach, which we refer to as time-dependent PCA.

Multiscale aspects of molecular motions: From molecular vibrations, conformational changes of biomolecules to cellular dynamics

Molecular aspects of living systems are important because it is the most basic aspects of life as exemplified in biochemistry and structural biology. Since molecules move due to interactive forces between atoms, physics plays an important role to understand the dynamic phenomena of living systems. Here we review our multiscale approaches for computationally treating different levels of molecular motions: vibrational dynamics of molecules, conformational change of biomolecules, and cellular dynamics using statistical-mechanics-based models.

Structural Biology and Molecular Modeling to Analyze the Entry of Bacterial Toxins and Virulence Factors into Host Cells

Understanding the functions and mechanisms of biological systems is an outstanding challenge. One way to overcome it is to combine together several approaches such as molecular modeling and experimental structural biology techniques. Indeed, the interplay between structural and dynamical properties of the system is crucial to unravel the function of molecular machinery’s. In this review, we focus on how molecular simulations along with structural information can aid in interpreting biological data. Here, we examine two different cases: (i) the endosomal translocation toxins (diphtheria, tetanus, botulinum toxins) and (ii) the activation of adenylyl cyclase inside the cytoplasm (edema factor, CyA, ExoY).

Pharmacological and molecular dynamics analyses of differences in inhibitor binding to human and nematode PDE4: Implications for management of parasitic nematodes

Novel chemical controls are needed that selectively target human, animal, and plant parasitic nematodes with reduced adverse effects on the host or the environment. We hypothesize that the phosphodiesterase (PDE) enzyme family represents a potential target for development of novel nematicides and anthelmintics. To test this, we identified six PDE families present in the nematode phylum that are orthologous to six of the eleven human PDE families. We characterized the binding interactions of family-selective PDE inhibitors with human and C. elegans PDE4 in conjunction with molecular dynamics (MD) simulations to evaluate differences in binding interactions of these inhibitors within the PDE4 catalytic domain. We observed that roflumilast (human PDE4-selective inhibitor) and zardaverine (selective for human PDE3 and PDE4) were 159- and 77-fold less potent, respectively, in inhibiting C. elegans PDE4. The pan-specific PDE inhibitor isobutyl methyl xanthine (IBMX) had similar affinity for nematode and human PDE4. Of 32 residues within 5 Å of the ligand binding site, five revealed significant differences in non-bonded interaction energies (van der Waals and electrostatic interaction energies) that could account for the differential binding affinities of roflumilast and zardaverine. One site (Phe506 in the human PDE4D3 amino acid sequence corresponding to Tyr253 in C. elegans PDE4) is predicted to alter the binding conformation of roflumilast and zardaverine (but not IBMX) into a less energetically favorable state for the nematode enzyme. The pharmacological differences in sensitivity to PDE4 inhibitors in conjunction with differences in the amino acids comprising the inhibitor binding sites of human and C. elegans PDE4 catalytic domains together support the feasibility of designing the next generation of anthelmintics/nematicides that could selectively bind to nematode PDEs.

In Silico Studies of Small Molecule Interactions with Enzymes Reveal Aspects of Catalytic Function

Small molecules, such as solvent, substrate, and cofactor molecules, are key players in enzyme catalysis. Computational methods are powerful tools for exploring the dynamics and thermodynamics of these small molecules as they participate in or contribute to enzymatic processes. In-depth knowledge of how small molecule interactions and dynamics influence protein conformational dynamics and function is critical for progress in the field of enzyme catalysis. Although numerous computational studies have focused on enzyme-substrate complexes to gain insight into catalytic mechanisms, transition states and reaction rates, the dynamics of solvents, substrates, and cofactors are generally less well studied. Also, solvent dynamics within the biomolecular solvation layer play an important part in enzyme catalysis, but a full understanding of its role is hampered by its complexity. Moreover, passive substrate transport has been identified in certain enzymes, and the underlying principles of molecular recognition are an area of active investigation. Enzymes are highly dynamic entities that undergo different conformational changes, which range from side chain rearrangement of a residue to larger-scale conformational dynamics involving domains. These events may happen nearby or far away from the catalytic site, and may occur on different time scales, yet many are related to biological and catalytic function. Computational studies, primarily molecular dynamics (MD) simulations, provide atomistic-level insight and site-specific information on small molecule interactions, and their role in conformational pre-reorganization and dynamics in enzyme catalysis. The review is focused on MD simulation studies of small molecule interactions and dynamics to characterize and comprehend protein dynamics and function in catalyzed reactions. Experimental and theoretical methods available to complement and expand insight from MD simulations are discussed briefly.

Molecular Modeling of the Catalytic Domain of CyaA Deepened the Knowledge of Its Functional Dynamics

Although CyaA has been studied for over three decades and revealed itself to be a very good prototype for developing various biotechnological applications, only a little is known about its functional dynamics and about the conformational landscape of this protein. Molecular dynamics simulations helped to clarify the view on these points in the following way. First, the model of interaction between AC and calmodulin (CaM) has evolved from an interaction centered on the surface between C-CaM hydrophobic patch and the α helix H of AC, to a more balanced view, in which the C-terminal tail of AC along with the C-CaM Calcium loops play an important role. This role has been confirmed by the reduction of the affinity of AC for calmodulin in the presence of R338, D360 and N347 mutations. In addition, enhanced sampling studies have permitted to propose a representation of the conformational space for the isolated AC. It remains to refine this representation using structural low resolution information measured on the inactive state of AC. Finally, due to a virtual screening study on another adenyl cyclase from Bacillus anthracis, weak inhibitors of AC have been discovered.

The allosteric switching mechanism in bacteriophage MS2

We use all-atom simulations to elucidate the mechanisms underlying conformational switching and allostery within the coat protein of the bacteriophage MS2. Assembly of most icosahedral virus capsids requires that the capsid protein adopts different conformations at precise locations within the capsid. It has been shown that a 19 nucleotide stem loop (TR) from the MS2 genome acts as an allosteric effector, guiding conformational switching of the coat protein during capsid assembly. Since the principal conformational changes occur far from the TR binding site, it is important to understand the molecular mechanism underlying this allosteric communication. To this end, we use all-atom simulations with explicit water combined with a path sampling technique to sample the MS2 coat protein conformational transition, in the presence and absence of TR-binding. The calculations find that TR binding strongly alters the transition free energy profile, leading to a switch in the favored conformation. We discuss changes in molecular interactions responsible for this shift. We then identify networks of amino acids with correlated motions to reveal the mechanism by which effects of TR binding span the protein. We find that TR binding strongly affects residues located at the 5-fold and quasi-sixfold interfaces in the assembled capsid, suggesting a mechanism by which the TR binding could direct formation of the native capsid geometry. The analysis predicts amino acids whose substitution by mutagenesis could alter populations of the conformational substates or their transition rates.

Flap Dynamics in Aspartic Proteases: A Computational Perspective

Recent advances in biochemistry and drug design have placed proteases as one of the critical target groups for developing novel small-molecule inhibitors. Among all proteases, aspartic proteases have gained significant attention due to their role in HIV/AIDS, malaria, Alzheimer's disease, etc. The binding cleft is covered by one or two β-hairpins (flaps) which need to be opened before a ligand can bind. After binding, the flaps close to retain the ligand in the active site. Development of computational tools has improved our understanding of flap dynamics and its role in ligand recognition. In the past decade, several computational approaches, for example molecular dynamics (MD) simulations, coarse-grained simulations, replica-exchange molecular dynamics (REMD) and metadynamics, have been used to understand flap dynamics and conformational motions associated with flap movements. This review is intended to summarize the computational progress towards understanding the flap dynamics of proteases and to be a reference for future studies in this field.

Influence of Solvation and Dynamics on the Mechanism and Kinetics of Nucleophilic Aromatic Substitution Reactions in Liquid Ammonia

The role of the solvent and the influence of dynamics on the kinetics and mechanism of the SNAr reaction of several halonitrobenzenes in liquid ammonia, using both static calculations and dynamic ab initio molecular dynamics simulations, are investigated. A combination of metadynamics and committor analysis methods reveals how this reaction can change from a concerted, one-step mechanism in gas phase to a stepwise pathway, involving a metastable Meisenheimer complex, in liquid ammonia. This clearly establishes, among others, the important role of the solvent and highlights the fact that accurately treating solvation is of crucial importance to correctly unravel the reaction mechanism. It is indeed shown that H-bond formation of the reacting NH3 with the solvent drastically reduces the barrier of NH3 addition. The halide elimination step, however, is greatly facilitated by proton transfer from the reacting NH3 to the solvent. Furthermore, the free energy surface strongly depends on the halide substituent and the number of electron-withdrawing nitro substituents.