Journal of Molecular Modeling最新文献

Context: In the realm of quantum chemistry, the accurate prediction of electronic structure and properties of nanostructures remains a formidable challenge. Density functional theory (DFT) and density matrix renormalization group (DMRG) have emerged as two powerful computational methods for addressing electronic correlation effects in diverse molecular systems. We compare ground-state energies ( $e_0$ ), density profiles ( $n$ ), and average entanglement entropies ( $\overline{S}$ ) in metals, insulators and at the transition from metal to insulator, in homogeneous, superlattices, and harmonically confined chains described by the fermionic one-dimensional Hubbard model. While for the homogeneous systems, there is a clear hierarchy between the deviations, $D\%\left(\overline{S}\right)<D\%\left(e_0\right)<\overline{D}\%\left(n\right)$ , and all the deviations decrease with the chain size; for superlattices and harmonic confinement, the relation among the deviations is less trivial and strongly dependent on the superlattice structure and the confinement strength considered. For the superlattices, in general, increasing the number of impurities in the unit cell represents lower precision in the DFT calculations. For the confined chains, DFT performs better for metallic phases, while the highest deviations appear for the Mott and band-insulator phases. This work provides a comprehensive comparative analysis of these methodologies, shedding light on their respective strengths, limitations, and applications.

Methods: The DFT calculations were performed using the standard Kohn-Sham scheme within the BALDA approach. It integrated the numerical Bethe-Ansatz (BA) solution of the Hubbard model as the homogeneous density functional within a local-density approximation (LDA) for the exchange-correlation energy. The DMRG algorithms were implemented using the ITensor library, which is based on the matrix product states (MPS) ansatz. The calculations were performed until the energy reaches convergence of at least ${10}^{-8}$ .

Context: Molecularly imprinted polymers (MIPs) have promising applications as synthetic antibodies for protein and peptide recognition. A critical aspect of MIP design is the selection of functional monomers and their adequate proportions to achieve materials with high recognition capacity toward their targets. To contribute to this goal, we calibrated a molecular dynamics protocol to reproduce the experimental trends in peptide recognition of 13 pre-polymerization mixtures reported in the literature for the peptide toxin melittin.

Methods: Three simulation conditions were tested for each mixture by changing the box size and the number of monomers and cross-linkers surrounding the template in a solvent-explicit environment. Fully atomistic MD simulations of 350 ns were conducted with the AMBER20 software, with ff19SB parameters for the peptide, gaff2 parameters for the monomers and cross-linkers, and the OPC water model. Template-monomer interaction energies under the LIE approach showed significant differences between high-affinity and low-affinity mixtures. Simulation systems containing 100 monomers plus cross-linkers in a cubic box of 90 ų successfully ranked the mixtures according to their experimental performance. Systems with higher monomer densities resulted in non-specific intermolecular contacts that could not account for the experimental trends in melittin recognition. The mixture with the best recognition capacity showed preferential binding to the 13-26-α-helix, suggesting a relevant role for this segment in melittin imprinting and recognition. Our findings provide insightful information to assist the computational design of molecularly imprinted materials with a validated protocol that can be easily extended to other templates.

Context: Geometrical knots are rare structural arrangements in proteins in which the polypeptide chain ties itself into a knot, which is very intriguing due to the uncertainty of their impact on the protein properties. Presently, classical molecular dynamics is the most employed technique in the few studies found on this topic, so any information on how the presence of knots affects the reactivity and electronic properties of proteins is even scarcer. Using the electronic structure methods and quantum chemical descriptors analysis, we found that the same amino-acid residues in the knot core have statistically larger values for the unknotted protein, for both hard-hard and soft-soft interaction descriptors. In addition, we present a computationally feasible protocol, where we show it is possible to separate the contribution of the geometrical knot to the reactivity and other electronic structure properties.

Methods: In order to investigate these systems, we used PRIMoRDiA, a new software developed by our research group, to explore the electronic structure of biological macromolecules. We evaluated several local quantum chemical descriptors to unveil relevant patterns potentially originating from the presence of the geometrical knot in two proteins, belonging to the ornithine transcarbamylase family. We compared several sampled structures from these two enzymes that are highly similar in both tertiary structure and function, but one of them has a knot whereas the other does not. The sampling was carried out through molecular dynamics simulations using ff14SB force field along 50 ns, and the semiempirical convergence was performed with PM7 Hamiltonian.

Context: Accurately predicting plasma protein binding rate (PPBR) and oral bioavailability (OBA) helps to better reveal the absorption and distribution of drugs in the human body and subsequent drug design. Although machine learning models have achieved good results in prediction accuracy, they often suffer from insufficient accuracy when dealing with data with irregular topological structures.

Methods: In view of this, this study proposes a pharmacokinetic parameter prediction framework based on graph convolutional networks (GCN), which predicts the PPBR and OBA of small molecule drugs. In the framework, GCN is first used to extract spatial feature information on the topological structure of drug molecules, in order to better learn node features and association information between nodes. Then, based on the principle of drug similarity, this study calculates the similarity between small molecule drugs, selects different thresholds to construct datasets, and establishes a prediction model centered on the GCN algorithm. The experimental results show that compared with traditional machine learning prediction models, the prediction model constructed based on the GCN method performs best on PPBR and OBA datasets with an inter-molecular similarity threshold of 0.25, with MAE of 0.155 and 0.167, respectively. In addition, in order to further improve the accuracy of the prediction model, GCN is combined with other algorithms. Compared to using a single GCN method, the distribution of the predicted values obtained by the combined model is highly consistent with the true values. In summary, this work provides a new method for improving the rate of early drug screening in the future.

Context: The debate over whether kinetic energy (KE) or potential energy (PE) are the fundamental energy components that contribute to forming covalent bonds has been enduring and stimulating over time. However, the supremacy of these energy components in reactions where multiple bonds are simultaneously formed or broken has yet to be explored. In this study, we use the reaction electronic flux (REF), an effective tool for investigating changes in driving electronic activity when bond formation or dissociation occurs in a chemical reaction, to examine the fluctuations in the KE and PE in a multi-bond reaction. To that end, the activation of CO ${}_2$ by low-valent group 14 catalysts through a concerted $\sigma$ -bond metathesis mechanism is analyzed. The findings of this preliminary study suggest that the REF can be utilized as a tool to rationalize alterations in the KE and PE in a multi-bond reaction. Specifically, analyses across the reaction coordinate reveal that changes in the KE and PE precede activation in the REF, stimulating the electronic activity where bond formation or dissociation processes dominate.

Methods: The activation of CO ${}_2$ by the low-valent LEH catalysts (L = N,N'-bis(2,6-diisopropyl phenyl)- $\beta$ -diketiminate; E = Si, Ge, Sn, and Pb) was studied along the reaction coordinate at the M06-2X/6-31 G(d,p)-LANL2DZ(E) level of theory. The respective minimum energy path calculations were obtained using the intrinsic reaction coordinate (IRC) procedure. The reaction electronic flux (REF) was calculated through the computation of the electronic chemical potential using the frontier molecular orbital approximation. Mayer bond orders along the reaction coordinate have been determined using the NBO 3.1 program in Gaussian16. Most of the reaction coordinate quantities reported in this study (REF, KE, PE, among others) have been determined using the Kudi program and custom Python scripts.

Context: The capacities of V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages to catalyze the ORR processes have been investigated. The acceptable pathways of ORR processes on V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages have been examined by DSD-PBEPBE-D3/aug-cc-pVDZ, PW91PW91/aug-cc-pVDZ, and COSMO model in the gas phase and water. The ΔG_reaction values of reaction steps of ORR pathways on V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages are calculated. The E_adoption and E_formation of V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages are negative values and these nanostructures are stable materials. The H₂O has the lowest E_adsorption on V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages. The *OH formation, creation of *OH-OH*, and formation of O* are rate-determining steps of ORR mechanisms. The overpotential values of ORR processes on V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages are 0.41, 0.37, and 0.30 V, respectively. The V-Al₃₆N₃₆ nanocage have lower overpotential for ORR processes than V-Si₇₂ and V-C₇₂ nanocages by DSD-PBEPBE-D3/aug-cc-pVDZ, PW91PW91/aug-cc-pVDZ, and COSMO model in the gas phase and water. The V-Al₃₆N₃₆ nanocage have more negative ∆G_reaction for reaction steps of ORR than V-Si₇₂ and V-C₇₂ nanocages. The V-Al₃₆N₃₆ nanocage with lower overpotential is proposed as an effective catalyst for ORR processes via studied pathways.

Methods: The DSD-PBEPBE-D3/aug-cc-pVDZ method has been used to optimize and calculate the frequencies of V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages in GAMESS software. The complexes of O, OH, OOH, and H₂O with V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages are optimized and frequencies are determined by the DSD-PBEPBE-D3/aug-cc-pVDZ method. The G_activation and ∆G_reaction of ORR pathways on V-Si₇₂, V-C₇₂, and V-Al₃₆N₃₆ nanocages are calculated by DSD-PBEPBE-D3/aug-cc-pVDZ, PW91PW91/aug-cc-pVDZ, and COSMO model in the gas phase and water.

Context: Multiwalled carbon nanotubes (MWCNTs) functionalized with lysine via 1,3-dipolar cycloaddition and conjugated to galactose or mannose are potential nanocarriers that can effectively bind to the lectin receptor in MDA-MB-231 or MCF-7 breast cancer cells. In this work, a method based on molecular dynamics (MD) simulation was used to predict the interaction of these functionalized MWCNTs with doxorubicin and obtain structural evidence that allows a better understanding of the drug loading and release process. The MD simulations showed that while doxorubicin only interacted with pristine MWCNTs through π-π stacking interactions, functionalized MWCNTs were also able to establish hydrogen bonds, suggesting that the functionalized groups improve doxorubicin loading. Moreover, the elevated adsorption levels observed for functionalized nanotubes further support this enhancement in loading efficiency. MD simulations also shed light on the intratumoral pH-specific release of doxorubicin from functionalized MWCNTs, which is induced by protonation of the daunosamine moiety. The simulations show that this change in protonation leads to a lower absorption of doxorubicin to the MWCNTs. The MD studies were then experimentally validated, where functionalized MWCNTs showed improved dispersion in aqueous medium compared to pristine MWCNTs and, in agreement with the computational predictions, increased drug loading capacity. Doxorubicin-loaded functionalized MWCNTs demonstrated specific release of doxorubicin in tumor microenvironment (pH = 5.0) with negligible release in the physiological pH (pH = 7.4). Furthermore, doxorubicin-free MWNCT nanoformulations exhibited insignificant cytotoxicity. The experimental studies yielded nearly identical results to the MD studies, underlining the usefulness of the method. Our functionalized MWCNTs represent promising non-toxic nanoplatforms with enhanced aqueous dispersibility and the potential for conjugation with ligands for targeted delivery of anti-cancer drugs to breast cancer cells.

Methods: The computational model of a pristine carbon nanotube was created with the buildCstruct 1.2 Python script. The lysinated functionalized groups were added with PyMOL and VMD. The carbon nanotubes and doxorubicin molecules were parameterized using the general AMBER force field, and RESP charges were determined using Gaussian 09. Molecular dynamics simulations were carried out with the AMBER 20 software package. Adsorption levels were calculated using the water-shell function of cpptraj. Cytotoxicity was evaluated via a MTT assay using MDA-MB-231 and MCF-7 breast cancer cells. Drug uptake of doxorubicin and doxorubicin-loaded MWCNTs was measured by fluorescence microscopy.

Context: Diabetes mellitus (DM) is a metabolic disorder disease that causes hyperglycemia conditions and associated with various chronic complications leading to mortality. Due to high toxicity of conventional diabetic drugs, the exploration of natural compounds as alternative diabetes treatments has been widely carried out. Previous in silico studies have highlighted berberine, a natural compound, as a promising alternative in antidiabetic therapy, potentially acting through various pathways, including the inhibition of the FOXO1 transcription factor in the gluconeogenesis pathway. However, the specific mechanism by which berberine interacts with FOXO1 remains unclear, and research in this area is relatively limited. Therefore, this study aims to determine the stability of berberine structure with FOXO1 based on RMSD, RMSF, binding energy, and trajectory analysis to determine the potential of berberine to inhibit the gluconeogenesis pathway. This research was conducted by in silico method with molecular docking using AutoDock4.2 and molecular dynamics study using Amber20, then visualized by VMD.

Methods: Docking between ligand and FOXO1 receptor was carried out with Autodock4.2. For molecular dynamics simulations, the force fields of DNA.OL15, protein.ff14SB, gaff2, and tip3p were used.

Context: The paper considers the features of the structure and dipole moments of several amino acids and their dipeptides which play an important role in the formation of the peptide nanotubes based on them. The influence of the features of their chirality (left L and right D) and the alpha-helix conformations of amino acids are taken into account. In particular, amino acids with aromatic rings, such as phenylalanine (Phe/F), and branched-chain amino acids (BCAAs)-leucine (Leu/L) and isoleucine (Ile/I)-as well as corresponding dipeptides (diphenylalanine (FF), dileucine (LL), and diisoleucine (II)) are considered. The main features and properties of these dipeptide structures and peptide nanotubes (PNTs), based on them, are investigated using computational molecular modeling and quantum-chemical semi-empirical calculations. Their polar, piezoelectric, and photoelectronic properties and features are studied in detail. The results of calculations of dipole moments and polarization, as well as piezoelectric coefficients and band gap width, for different types of helical peptide nanotubes are presented. The calculated values of the chirality indices of various nanotubes are given, depending on the chirality of the initial dipeptides-the results obtained are consistent with the law of changes in the type of chirality as the hierarchy of molecular structures becomes more complex. The influence of water molecules in the internal cavity of nanotubes on their physical properties is estimated. A comparison of the results of these calculations by various computational methods with the available experimental data is presented and discussed.

Method: The main tool for molecular modeling of all studied nanostructures in this work was the HyperChem 8.01 software package. The main approach used here is the Hartree-Fock (HF) self-consistent field (SCF) with various quantum-chemical semi-empirical methods (AM1, PM3, RM1) in the restricted Hartree-Fock (RHF) and in the unrestricted Hartree-Fock (UHF) approximations. Optimization of molecular systems and the search for their optimal geometry is carried out in this work using the Polak-Ribeire algorithm (conjugate gradient method), which determines the optimized geometry at the point of their minimum total energy. For such optimized structures, dipole moments D and electronic energy levels (such as E_HOMO and E_LUMO), as well as the band gap E_g = E_LUMO - E_HOMO, were then calculated. For each optimized molecular structure, the volume was calculated using the QSAR program implemented also in the HyperChem software package.

Context: 1,3-Dithiole-2-thione-4,5-dithiolate (dmit) ligands are known for their conductive and optical properties. Dmit compounds have been assessed for use in sensor devices, information storage, spintronics, and optical material applications. Associations with various metallic centers endow dmit complexes with magnetic, optical, conductive, and antioxidant properties. Optical doping can facilitate the fabrication of magnetic conductor materials from ground-state nonmagnetic cations. While most studied complexes involve transition-metal centers due to their diverse chemistry, compounds with representative elements are less explored in the literature. This study investigated the structural and electronic properties of bisdmit complexes with representative Bi(III), Sb(III), and Zn(II) cations. AIMD calculations revealed two new geometries for Bi(III) and Zn(II) complexes, diverging from the isolated geometry typically used in quantum chemical calculations. The coordination of acetonitrile molecules to the cationic centers of the complexes resulted in unstable structures, while the dimerization of the complexes was stable. SA-CASSCF/NEVPT2 calculations were applied to the structures of the isolated complexes and stable dimers, confirming the multireference character of the electronic structure of the three systems and the multiconfigurational character of the Bi(III) complex. The electronic spectra simulated by the STEOM-DLPNO-CCSD calculations accurately reproduced the experimental UV‒Vis spectra indicating the participation of the isolated Bi(III) dmit complex and its dimeric form in solution.

Methodology: AIMD calculations of the dmit salts were conducted using the GFN2-xTB method with 60 explicit acetonitrile molecules as the solvent at 300 K for a total simulation time of 50.0 ps, with printing intervals of 0.5 fs. The final geometries were optimized employing the PBEh-3c compound method, incorporating implicit conductor-like polarizable continuum model (CPCM) solvation for acetonitrile. Local energy decomposition (LED) analysis at the DLPNO-CCSD(T)/Def2-TZVP level of theory was utilized to investigate the stability of the complex geometries identified by AIMD. The electronic structures of the complexes were assessed using the SA-CASSCF/NEVPT2/Def2-TZVP method to confirm the multiconfigurational and multireference nature of their electronic structures. Electronic spectra were analyzed using the STEOM-DLPNO-CCSD/Def2-TZVP method, with CPCM used to simulate an acetonitrile medium.