Finite-size effects in the static structure factor S(k) and S(0) for a two-dimensional Yukawa liquid.

Finite-size effects in the static structure factor S(k) are analyzed for an amorphous substance. As the number of particles is reduced, S(0) increases greatly, up to an order of magnitude. Meanwhile, there is a decrease in the height of the first peak S_{peak}. These finite-size effects are modeled accurately by the Binder formula for S(0) and our empirical formula for S_{peak}. Procedures are suggested to correct for finite-size effects in S(k) data and in the hyperuniformity index H≡S(0)/S_{peak}. These principles generally apply to S(k) obtained from particle positions in noncrystalline substances. The amorphous substance we simulate is a two-dimensional liquid, with a soft Yukawa interaction modeling a dusty plasma experiment.

Low-THz Vibrations of Biological Membranes.

A growing body of work has linked key biological activities to the mechanical properties of cellular membranes, and as a means of identification. Here, we present a computational approach to simulate and compare the vibrational spectra in the low-THz region for mammalian and bacterial membranes, investigating the effect of membrane asymmetry and composition, as well as the conserved frequencies of a specific cell. We find that asymmetry does not impact the vibrational spectra, and the impact of sterols depends on the mobility of the components of the membrane. We demonstrate that vibrational spectra can be used to distinguish between membranes and, therefore, could be used in identification of different organisms. The method presented, here, can be immediately extended to other biological structures (e.g., amyloid fibers, polysaccharides, and protein-ligand structures) in order to fingerprint and understand vibrations of numerous biologically-relevant nanoscale structures.

Domain-agnostic predictions of nanoscale interactions in proteins and nanoparticles.

Although challenging, the accurate and rapid prediction of nanoscale interactions has broad applications for numerous biological processes and material properties. While several models have been developed to predict the interaction of specific biological components, they use system-specific information that hinders their application to more general materials. Here we present NeCLAS, a general and efficient machine learning pipeline that predicts the location of nanoscale interactions, providing human-intelligible predictions. NeCLAS outperforms current nanoscale prediction models for generic nanoparticles up to 10-20 nm, reproducing interactions for biological and non-biological systems. Two aspects contribute to these results: a low-dimensional representation of nanoparticles and molecules (to reduce the effect of data uncertainty), and environmental features (to encode the physicochemical neighborhood at multiple scales). This framework has several applications, from basic research to rapid prototyping and design in nanobiotechnology.

Comparison of the static structure factor at long wavelengths for a dusty plasma liquid and other liquids.

Especially small values of the static structure factor S(k) at long wavelengths, i.e., small k, were obtained in an analysis of experimental data, for a two-dimensional dusty plasma in its liquid state. For comparison, an analysis of S(k) data was carried out for many previously published experiments with other liquids. The latter analysis indicates that the magnitude of S(k) at small k is typically in a range 0.02-0.13. In contrast, the corresponding value for a dusty plasma liquid was found to be as small as 0.0139. Another basic finding for the dusty plasma liquid is that S(k) at small k generally increases with temperature, with its lowest value, noted above, occurring near the melting point. Simulations were carried out for the dusty plasma liquid, and their results are generally consistent with the experiment. Since a dusty plasma has a soft interparticle interaction, our findings support earlier theoretical suggestions that a useful design strategy for creating materials having exceptionally low values of S(0), so-called hyperuniform materials, is the use of a condensed material composed of particles that interact softly at their periphery.

Chiral chromatography and surface chirality of carbon nanoparticles.

Chiral carbon nanoparticles (CNPs) represent a rapidly evolving area of research for optical and biomedical technologies. Similar to small molecules, applications of CNPs as well as fundamental relationships between their optical activity and structural asymmetry would greatly benefit from their enantioselective separations by chromatography. However, this technique remains in its infancy for chiral carbon and other nanoparticles. The possibility of effective separations using high performance liquid chromatography (HPLC) with chiral stationary phases remains an open question whose answer can also shed light on the components of multiscale chirality of the nanoparticles. Herein, we report a detailed methodology of HPLC for successful separation of chiral CNPs and establish a path for its future optimization. A mobile phase of water/acetonitrile was able to achieve chiral separation of CNPs derived from L- and D-cysteine denoted as L-CNPs and D-CNPs. Molecular dynamics simulations show that the teicoplanin-based stationary phase has a higher affinity for L-CNPs than for D-CNPs, in agreement with experiments. The experimental and computational findings jointly indicate that chiral centers of chiral CNPs are present at their surface, which is essential for the multiple applications of these chiral nanostructures and equally essential for interactions with biomolecules and circularly polarized photons.

Struct2Graph: a graph attention network for structure based predictions of protein-protein interactions.

BACKGROUND: Development of new methods for analysis of protein-protein interactions (PPIs) at molecular and nanometer scales gives insights into intracellular signaling pathways and will improve understanding of protein functions, as well as other nanoscale structures of biological and abiological origins. Recent advances in computational tools, particularly the ones involving modern deep learning algorithms, have been shown to complement experimental approaches for describing and rationalizing PPIs. However, most of the existing works on PPI predictions use protein-sequence information, and thus have difficulties in accounting for the three-dimensional organization of the protein chains. RESULTS: In this study, we address this problem and describe a PPI analysis based on a graph attention network, named Struct2Graph, for identifying PPIs directly from the structural data of folded protein globules. Our method is capable of predicting the PPI with an accuracy of 98.89% on the balanced set consisting of an equal number of positive and negative pairs. On the unbalanced set with the ratio of 1:10 between positive and negative pairs, Struct2Graph achieves a fivefold cross validation average accuracy of 99.42%. Moreover, Struct2Graph can potentially identify residues that likely contribute to the formation of the protein-protein complex. The identification of important residues is tested for two different interaction types: (a) Proteins with multiple ligands competing for the same binding area, (b) Dynamic protein-protein adhesion interaction. Struct2Graph identifies interacting residues with 30% sensitivity, 89% specificity, and 87% accuracy. CONCLUSIONS: In this manuscript, we address the problem of prediction of PPIs using a first of its kind, 3D-structure-based graph attention network (code available at https://github.com/baranwa2/Struct2Graph ). Furthermore, the novel mutual attention mechanism provides insights into likely interaction sites through its unsupervised knowledge selection process. This study demonstrates that a relatively low-dimensional feature embedding learned from graph structures of individual proteins outperforms other modern machine learning classifiers based on global protein features. In addition, through the analysis of single amino acid variations, the attention mechanism shows preference for disease-causing residue variations over benign polymorphisms, demonstrating that it is not limited to interface residues.

Distance-dependent resonance energy transfer in alkyl-terminated Si nanocrystal solids.

Understanding and controlling the energy transfer between silicon nanocrystals is of significant importance for the design of efficient optoelectronic devices. However, previous studies on silicon nanocrystal energy transfer were limited because of the strict requirements to precisely control the inter-dot distance and to perform all measurements in air-free environments to preclude the effect of ambient oxygen. Here, we systematically investigate the distance-dependent resonance energy transfer in alkyl-terminated silicon nanocrystals for the first time. Silicon nanocrystal solids with inter-dot distances varying from 3 to 5 nm are fabricated by varying the length and surface coverage of alkyl ligands in solution-phase and gas-phase functionalized silicon nanocrystals. The inter-dot energy transfer rates are extracted from steady-state and time-resolved photoluminescence measurements, enabling a direct comparison to theoretical predictions. Our results reveal that the distance-dependent energy transfer rates in Si NCs decay faster than predicted by the Förster mechanism, suggesting higher-order multipole interactions.

Stochastic and network analysis of polycyclic aromatic growth in a coflow diffusion flame.

An important step in predicting the growth of soot nanoparticles is understanding how gas phase variations affect the formation of their aromatic precursors. Once formed, these aromatic structures begin to assemble into nanoparticles and, regardless of the clustering process, the molecular properties of the aromatic precursors play an important role. Leveraging existing experimental data collected from a coflow Jet A-1 surrogate diffusion flame, in this paper we report on a detailed study of the spatial evolution of molecular structures of polycyclic aromatic compounds (PACs) and their corresponding formation pathways. To this end, we employed the SNapS2 kinetic Monte Carlo software to simulate the chemical evolution of PACs along multiple streamlines. The results show that growth only occurs along streamlines that traverse regions of high acetylene concentrations in the center of the flame. The PACs predicted in various conditions show diverse chemical properties, including aliphatic chains, five-membered, and heteroaromatic rings. PACs in streamlines close to the flame wings begin growing immediately due to the high temperature and large amounts of radical species, while PACs originating along inner streamlines do not appreciably grow until they pass through an area characterized by high radical concentrations. Using graph theory and network analysis, we investigated the complex reaction network generated by SNapS2 and determined that the growth pathways of many PACs center around a few stable structures that also promote oxygen addition reactions due to their morphology and long lifetimes. These pathways play a more significant role along streamlines near the centerline, compared to the flame wings, which show more variety due to the highly reactive environment encountered during early growth. The results of this study provide insights on the reaction pathways that determine the properties of PACs at different flame locations as well as information on the chemical characteristics of the formed PACs, with emphasis on oxygenated structures.

On Drug-Membrane Permeability of Antivirals for SARS-CoV-2.

One of the key parameters required to identify effective drugs is membrane permeability, as a compound intended for an intracellular target with poor permeability will have low efficacy. In this paper, we leverage a computational approach recently developed by our group to study the interactions between nanoparticles and mammalian membranes to study the time of entry of a variety of drugs into the viral envelope of coronavirus as well as cellular organelles. Using a combination of all-atoms molecular dynamics simulations and statistical analysis, we consider both drug characteristics and membrane properties to determine the behavior of 79 drugs and their interactions with the viral envelope, composed of the membrane and spike protein, as well as five other membranes that correspond to various mammalian compartments (lysosome, plasma, Golgi, mitochondrial, and endoplasmic reticulum membranes). The results highlight important trends that can be exploited for drug design, from the relatively high permeability of the viral envelope and the effect of transmembrane proteins, to the differences in permeability between organelles. When compared with bioavailability data present in the literature, the model results suggest a negative correlation between time of permeation and bioavailability of promising drugs. The method is general and flexible and can be employed for a variety of molecules, from small drugs to small nanoparticles, as well to a variety of biological membranes. Overall, the results indicate that this model can contribute to the identification of successful drugs as it predicts the ability of compounds to reach both intended and unintended intracellular targets.

A deep learning architecture for metabolic pathway prediction.

MOTIVATION: Understanding the mechanisms and structural mappings between molecules and pathway classes are critical for design of reaction predictors for synthesizing new molecules. This article studies the problem of prediction of classes of metabolic pathways (series of chemical reactions occurring within a cell) in which a given biochemical compound participates. We apply a hybrid machine learning approach consisting of graph convolutional networks used to extract molecular shape features as input to a random forest classifier. In contrast to previously applied machine learning methods for this problem, our framework automatically extracts relevant shape features directly from input SMILES representations, which are atom-bond specifications of chemical structures composing the molecules. RESULTS: Our method is capable of correctly predicting the respective metabolic pathway class of 95.16% of tested compounds, whereas competing methods only achieve an accuracy of 84.92% or less. Furthermore, our framework extends to the task of classification of compounds having mixed membership in multiple pathway classes. Our prediction accuracy for this multi-label task is 97.61%. We analyze the relative importance of various global physicochemical features to the pathway class prediction problem and show that simple linear/logistic regression models can predict the values of these global features from the shape features extracted using our framework. AVAILABILITY AND IMPLEMENTATION: https://github.com/baranwa2/MetabolicPathwayPrediction. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Predicting the Time of Entry of Nanoparticles in Lipid Membranes.

The number of engineered nanoparticles for applications in the biomedical arena has grown tremendously over the last years due to advances in the science of synthesis and characterization. For most applications, the crucial step is the transport through a physiological cellular membrane. However, the behavior of nanoparticles in a biological matrix is a very complex problem that depends not only on the type of nanoparticle but also on its size, shape, phase, surface charge, chemical composition, and agglomeration state. In this paper, we introduce a streamlined theoretical model that predicts the average time of entry of nanoparticles in lipid membranes, using a combination of molecular dynamics simulations and statistical approaches. The model identifies four parameters that separate the contributions of nanoparticle characteristics (i.e., size, shape, solubility) from the membrane properties (density distribution). This factorization allows the inclusion of data obtained from both experimental and computational sources, as well as a rapid estimation of large sets of permutations in membranes. The robustness of the model is supported by experimental data carried out in lipid vesicles encapsulating graphene quantum dots as nanoparticles. Given the high level of interest across multiple areas of study in modulating intracellular targets, and the need to understand and improve the applications of nanoparticles and to assess their effect on human health (i.e., cytotoxicity, bioavailability), this work contributes to the understanding and prediction of interactions between nanoparticles and lipid membranes.

Anti-Biofilm Activity of Graphene Quantum Dots via Self-Assembly with Bacterial Amyloid Proteins.

Bacterial biofilms represent an essential part of Earth's ecosystem that can cause multiple ecological, technological, and health problems. The environmental resilience and sophisticated organization of biofilms are enabled by the extracellular matrix that creates a protective network of biomolecules around the bacterial community. Current anti-biofilm agents can interfere with extracellular matrix production but, being based on small molecules, are degraded by bacteria and rapidly diffuse away from biofilms. Both factors severely reduce their efficacy, while their toxicity to higher organisms creates additional barriers to their practicality. In this paper, we report on the ability of graphene quantum dots to effectively disperse mature amyloid-rich Staphylococcus aureus biofilms, interfering with the self-assembly of amyloid fibers, a key structural component of the extracellular matrix. Mimicking peptide-binding biomolecules, graphene quantum dots form supramolecular complexes with phenol-soluble modulins, the peptide monomers of amyloid fibers. Experimental and computational results show that graphene quantum dots efficiently dock near the N-terminus of the peptide and change the secondary structure of phenol-soluble modulins, which disrupts their fibrillation and represents a strategy for mitigation of bacterial communities.

Critical Assessment of Photoionization Efficiency Measurements for Characterization of Soot-Precursor Species.

We present a critical evaluation of photoionization efficiency (PIE) measurements coupled with aerosol mass spectrometry for the identification of condensed soot-precursor species extracted from a premixed atmospheric-pressure ethylene/oxygen/nitrogen flame. Definitive identification of isomers by any means is complicated by the large number of potential isomers at masses likely to comprise particles at flame temperatures. This problem is compounded using PIE measurements by the similarity in ionization energies and PIE-curve shapes among many of these isomers. Nevertheless, PIE analysis can provide important chemical information. For example, our PIE curves show that neither pyrene nor fluoranthene alone can describe the signal from C16H10 isomers and that coronene alone cannot describe the PIE signal from C24H12 species. A linear combination of the reference PIE curves for pyrene and fluoranthene yields good agreement with flame-PIE curves measured at 202 u, which is consistent with pyrene and fluoranthene being the two major C16H10 isomers in the flame samples, but does not provide definite proof. The suggested ratio between fluoranthene and pyrene depends on the sampling conditions. We calculated the values of the adiabatic-ionization energy (AIE) of 24 C16H10 isomers. Despite the small number of isomers considered, the calculations show that the differences in AIEs between several of the isomers can be smaller than the average thermal energy at room temperature. The calculations also show that PIE analysis can sometimes be used to separate hydrocarbon species into those that contain mainly aromatic rings and those that contain significant aliphatic content for species sizes investigated in this study. Our calculations suggest an inverse relationship between AIE and the number of aromatic rings. We have demonstrated that further characterization of precursors can be facilitated by measurements that test species volatility.

Photoionization Efficiencies of Five Polycyclic Aromatic Hydrocarbons.

We have measured photoionization-efficiency curves for pyrene, fluoranthene, chrysene, perylene, and coronene in the photon energy range of 7.5-10.2 eV and derived their photoionization cross-section curves in this energy range. All measurements were performed using tunable vacuum ultraviolet (VUV) radiation generated at the Advanced Light Source synchrotron at Lawrence Berkeley National Laboratory. The VUV radiation was used for photoionization, and detection was performed using a time-of-flight mass spectrometer. We measured the photoionization efficiency of 2,5-dimethylfuran simultaneously with those of pyrene, fluoranthene, chrysene, perylene, and coronene to obtain references of the photon flux during each measurement from the known photoionization cross-section curve of 2,5-dimethylfuran.

Chiral Graphene Quantum Dots.

Chiral nanostructures from metals and semiconductors attract wide interest as components for polarization-enabled optoelectronic devices. Similarly to other fields of nanotechnology, graphene-based materials can greatly enrich physical and chemical phenomena associated with optical and electronic properties of chiral nanostructures and facilitate their applications in biology as well as other areas. Here, we report that covalent attachment of l/d-cysteine moieties to the edges of graphene quantum dots (GQDs) leads to their helical buckling due to chiral interactions at the "crowded" edges. Circular dichroism (CD) spectra of the GQDs revealed bands at ca. 210-220 and 250-265 nm that changed their signs for different chirality of the cysteine edge ligands. The high-energy chiroptical peaks at 210-220 nm correspond to the hybridized molecular orbitals involving the chiral center of amino acids and atoms of graphene edges. Diverse experimental and modeling data, including density functional theory calculations of CD spectra with probabilistic distribution of GQD isomers, indicate that the band at 250-265 nm originates from the three-dimensional twisting of the graphene sheet and can be attributed to the chiral excitonic transitions. The positive and negative low-energy CD bands correspond to the left and right helicity of GQDs, respectively. Exposure of liver HepG2 cells to L/D-GQDs reveals their general biocompatibility and a noticeable difference in the toxicity of the stereoisomers. Molecular dynamics simulations demonstrated that d-GQDs have a stronger tendency to accumulate within the cellular membrane than L-GQDs. Emergence of nanoscale chirality in GQDs decorated with biomolecules is expected to be a general stereochemical phenomenon for flexible sheets of nanomaterials.

Stochastic atomistic simulation of polycyclic aromatic hydrocarbon growth in combustion.

Nanoparticles formed in gas phase environments, such as combustion, have an important impact on society both as engineering components and hazardous pollutants. A new software package, the Stochastic Nanoparticle Simulator (SNAPS) was developed, applying a stochastic chemical kinetics methodology, to computationally investigate the growth of nanoparticle precursors through trajectories of chemical reactions. SNAPS was applied to characterize the growth of polycyclic aromatic hydrocarbons (PAHs), important precursors of carbonaceous nanoparticles and soot, in a premixed laminar benzene flame, using a concurrently developed PAH growth chemical reaction mechanism, as well as an existing benzene oxidation mechanism. Simulations of PAH ensembles successfully predicted existing experimentally measured data and provided novel insight into chemical composition and reaction pathways. The most commonly observed PAH isomers in simulations showed the importance of 5-membered rings, which contrasts with traditionally assumed compositions involving primarily pericondensed 6-membered rings. In addition, the chemical growth of PAHs involved complex sequences of highly reversible reactions, rather than relatively direct routes of additions and ring closures. Furthermore, the most common reactions involved 5-membered rings, suggesting their importance to PAH growth. The framework developed in this work will facilitate future investigation of particle inception and soot formation and will benefit engineering of novel combustion technologies to mitigate harmful emissions.

Free energy calculation of permeant-membrane interactions using molecular dynamics simulations.

Nanotoxicology, the science concerned with the safe use of nanotechnology and nanostructure design for biological applications, is a field of research that has recently received great attention, as a result of the rapid growth in nanotechnology. Many nanostructures are of a scale and chemical composition similar to many biomolecular environments, and recent papers have reported evident toxicity of selected nanoparticles. Molecular simulations can help develop a mechanistic understanding of how structural properties affect bioactivity. In this chapter, we describe how to compute the free energy of interactions between cellular membranes and benzene, the main constituent of some toxic carbonaceous particles, with well-tempered metadynamics. This algorithm reconstructs the free energy surface and accelerates rare events in a coarse-grained representation of the system.

Effect of molecular configuration on binary diffusion coefficients of linear alkanes.

Mass diffusion coefficients are critically related to the predictive capability of computational combustion modeling. To date, the most common approach used to determine the molecular transport of gases is the Boltzmann transport equation of the gas kinetic theory. The Chapman-Enskog (CE) solution of this transport equation, combined with Lennard-Jones potential parameters, suggests a simple analytical expression for computing self and mutual diffusion coefficients. This approach has been applied over a wide range of flame modeling conditions due to its minimal computational requirement, despite the fact that the theory was developed only for molecules that have a spherical structure. In this study, we computed the binary diffusion coefficients of linear alkanes using all-atom molecular dynamics simulations over the temperature range 500-1000 K. The effect of molecular configurations on diffusion coefficients was determined relating the radii of gyration of the molecules to their corresponding collision diameters. The comparison between diffusion coefficients determined with molecular dynamics and the values obtained from the CE theory shows significant discrepancies, especially for nonspherical molecules. This study reveals the inability of CE theory with spherical potentials to account for the effect of molecular shapes on diffusion coefficients.