Salient Structural Features Research Articles

Multi-granularity Cross Transformer Network for person re-identification

Person re-identification (Re-ID) aims to retrieve the same person in the gallery. Transformers have been introduced to the Re-ID task due to their excellent ability to model long-range dependency. However, due to the properties of the global attention mechanism, they are less effective in capturing the discriminative local semantics of pedestrians compared to convolutional operations. To address this issue, we present a Multi-granularity Cross Transformer Network (MCTN) that progressively learns salient features of different local structures in a global context. Specifically, we first utilize a Multi-granularity Convolutional Layer (MCL) to investigate salient pedestrian features at various granularities. On this basis, we propose a Pyramidal Cross Transformer learning layer (PCT), which contains a pyramidal division of pedestrian image feature maps, differentiated feature extraction of different parts of pedestrians, and cross attention to exploring the local–global relationship of the feature map. It allows effective mining of local information in the global structure from a coarse-to-fine perspective. Furthermore, to enhance the interaction between low-level detailed features and high-level semantic features, a Hierarchical Aggregation Strategy (HAS) is introduced to fuse features learned by cross attention learning at different stages. Pedestrian features learned in shallow layers will serve as global priors for semantics learning in deep layers. We evaluate our method on four large-scale Re-ID datasets, and the experimental results reveal that the proposed method outperforms the state-of-the-art methods.

Electronic structure of CeAuAl$_3$ using density functional theory

We studied the magnetic properties and electronic structure of CeAuAl_3CeAuAl3 using density functional theory. This compound shows a large Sommerfeld coefficient, a Kondo temperature, {T_K=4\,K}TK=4K and antiferromagnetic order below {T_N=1.1\,K}TN=1.1K. We calculated the magnetic groundstate of {CeAuAl_3}CeAuAl3 and the magnetic anisotropy energies. Treating the 4f-electrons as localized with DFT+U we obtain a good match with the magnetic properties observed experimentally. We also report salient features of the electronic structure of {CeAuAl_3}CeAuAl3, including features of the Fermi surface and associated quantum oscillatory spectra, when the 4f-electrons are treated either as localized or itinerant.

Downward Momentum Flux: An Important Mechanism of Typhoon Maintaining Ground Destructive Force

AbstractUnder the background of global climate change, typhoons have been attracting increasing attention due to their extraordinary destructive potential and great impact on the coastal areas of the South China Sea. Although the risk of strong winds related to typhoons has long been of interest, less is known about the underlying mechanism responsible for the severe near‐surface winds. Using eddy covariance data collected at four different heights on a 365‐m meteorological tower located in a coastal region, the characteristics of the convective typhoon boundary layer and the associated turbulence structures are compared with their counterparts in the “textbook” dynamically unstable boundary layer. In the convective typhoon boundary layer, bulk wind shear predominates the generation of mechanical turbulence, enhancing the vertical correlation between vertical layers. The spectral analysis highlights the salient features of turbulent structures under the convective typhoon boundary layer, confirming that the gust disturbance with the nondimensional frequency ranging from 0.003 to 0.3 modulates not only turbulent transports but also the horizontal flow. Such gusts with reduced phase difference enhance the downward momentum transport, mainly responsible for the maintenance of the strong near‐surface winds during typhoon landfalls.

Generalized Disappointment Aversion and the Variance Term Structure

Abstract Contrary to leading asset pricing theories, recent empirical evidence indicates that financial markets compensate only short-term equity variance risk. An equilibrium model with generalized disappointment aversion risk preferences and rare events reconciles salient features of the variance term structure. In addition, a calibration explains the variance and skew risk premiums in equity returns and the implied volatility skew of index options while capturing standard moments of fundamentals, equity returns, and the risk-free rate. The key intuition for the results stems from substantial countercyclical risk aversion induced by endogenous variation in the probability of disappointing events in consumption growth.

Insights Into Resveratrol as an Inhibitor Against Aβ1-42 Peptide Aggregation: A Molecular Dynamics Simulation Study

Background: Resveratrol (RSV), a polyphenolic compound, is reported to have anti-aggregation properties against Amyloid-beta peptides. It is, therefore, significant to understand the mechanism of inhibition of Aβ1-42 peptide aggregation by the RSV at the molecular level. We have used Molecular docking along with Molecular dynamics (MD) simulation techniques to address the role of RSV in the inhibition of Aβ1-42 pep-tide aggregation. Objective: To understand the role of Resveratrol on the Aβ1-42 peptide aggregation. Method: In this computational study, we have docked the RSV to Aβ1-42 peptide using Molecular Docking software and then performed MD simulation for the Aβ1-42 peptide monomer Aβ1-42 peptide-RSV complex using the AMBER force field. From the analysis of MD trajectories, we obtained salient structural features and determined the Binding Free Energy(BFE) and Per-residue Energy Decomposition Analysis (PRED) using MM-PBSA/GBSA method. Results: The secondary structure and the conformational analysis obtained from MD trajectories show that the binding of RSV with the Aβ1-42 peptide monomer causes an increase in the helical content in the structure of the Aβ1-42 peptide. The BFE and PRED results show a high binding affinity (GBtotal=-11.07 kcal mol-1; PBtotal= -1.82 kcal mol-1) of RSV with Aβ1-42 peptide. Also, we found the RSV to interact with cru-cial residues (Asp 23 and Lys 28) of the Aβ1-42 peptide. These residues play a signif-icant role in facilitating the formation of toxic amyloid oligomers and amyloid fibrils. The salt bridge interaction between these residues D23–K28 was found to be destabi-lized in the Aβ1-42 peptide when it is complexed with RSV. Conclusion: In summary, it can be concluded that Resveratrol greatly aids the preven-tion of Aβ1-42 peptide aggregation. Therefore, it can be considered a possible drug candidate for therapeutic strategies for Alzheimer’s disease.

BIRGU Net: deformable brain magnetic resonance image registration using gyral-net map and 3D Res-Unet

Deformable image registration is a fundamental procedure in medical imaging. Recently, deep learning-based deformable image registrations have achieved fast registration by learning the spatial correspondence from image pairs. However, it remains challenging in brain image registration due to the structural complexity of individual brains and the lack of ground truth for anatomical correspondences between the brain image pairs. This work devotes to achieving an end-to-end unsupervised brain deformable image registration method using the gyral-net map and 3D Res-Unet (BIRGU Net). Firstly, the gyral-net map was introduced to represent the 3D global cortex complex information of the brain image since it was considered as one of the anatomical landmarks, which can help to extract the salient structural feature of individual brains for registration. Secondly, the variant of 3D U-net architecture involving dual residual strategies was designed to map the image into the deformation field effectively and to prevent the gradient from vanishing as well. Finally, double regularized terms were imposed on the deformation field to guide the network training for leveraging the smoothness and the topology preservation of the deformation field. The registration procedure was trained in an unsupervised manner, which addressed the lack of ground truth for anatomical correspondences between the brain image pairs. The experimental results on four public data sets demonstrate that the extracted gyral-net can be an auxiliary feature for registration and the proposed network with the designed strategies can improve the registration performance since the Dice similarity coefficient (DSC) and normalized mutual information (NMI) are higher and the time consumption is comparable than the state-of-the-art. The code is available at https://github.com/mynameiswode/BIRGU-Net .

Targeting the Major Groove of the Palindromic d(GGCGCC)2 Sequence by Oligopeptide Derivatives of Anthraquinone Intercalators.

GC-rich sequences are recurring motifs in oncogenes and retroviruses and could be targeted by noncovalent major-groove therapeutic ligands. We considered the palindromic sequence d(G1G2C3G4C5C6)2, and designed several oligopeptide derivatives of the anticancer intercalator mitoxantrone. The stability of their complexes with an 18-mer oligonucleotide encompassing this sequence in its center was validated using polarizable molecular dynamics. We report the most salient structural features of two novel compounds, having a dialkylammonium group as a side chain on both arms. The anthraquinone ring is intercalated in the central d(CpG)2 sequence with its long axis perpendicular to that of the two base pairs. On each strand, this enables each ammonium group to bind in-register to O6/N7 of the two facing G bases upstream. We subsequently designed tris-intercalating derivatives, each dialkylammonium substituted with a connector to an N9-aminoacridine intercalator extending our target range from a six- to a ten-base-pair palindromic sequence, d(C1G2G3G4C5G6C7C8C9G10)2. The structural features of the complex of the most promising derivative are reported. The present design strategy paves the way for designing intercalator-oligopeptide derivatives with even higher selectivity, targeting an increased number of DNA bases, going beyond ten.

New MS lesion segmentation with deep residual attention gate U-Net utilizing 2D slices of 3D MR images.

Multiple sclerosis (MS) is an autoimmune disease that causes lesions in the central nervous system of humans due to demyelinating axons. Magnetic resonance imaging (MRI) is widely used for monitoring and measuring MS lesions. Automated methods for MS lesion segmentation have usually been performed on individual MRI scans. Recently, tracking lesion activity for quantifying and monitoring MS disease progression, especially detecting new lesions, has become an important biomarker. In this study, a unique pipeline with a deep neural network that combines U-Net, attention gate, and residual learning is proposed to perform better new MS lesion segmentation using baseline and follow-up 3D FLAIR MR images. The proposed network has a similar architecture to U-Net and is formed from residual units which facilitate the training of deep networks. Networks with fewer parameters are designed with better performance through the skip connections of U-Net and residual units, which facilitate information propagation without degradation. Attention gates also learn to focus on salient features of the target structures of various sizes and shapes. The MSSEG-2 dataset was used for training and testing the proposed pipeline, and the results were compared with those of other proposed pipelines of the challenge and experts who participated in the same challenge. According to the results over the testing set, the lesion-wise F1 and dice scores were obtained as a mean of 48 and 44.30%. For the no-lesion cases, the number of tested and volume of tested lesions were obtained as a mean of 0.148 and 1.488, respectively. The proposed pipeline outperformed 22 proposed pipelines and ranked 8th in the challenge.

Uptake of Am(III) Ions and Eu(III) Ions Using Cyclic Substituted N, O‐hybrid 1,10‐Phenanthroline Derived Phosphine Oxide Ligands ‐ A DFT Exploration

AbstractThe spent nuclear fuel has radioactive species such as actinides and lanthanides which hamper its reprocessing. This issue is undertaken by designing suitable organic ligands to strip off actinides from lanthanides. Due to the versatility of phenanthroline‐derived N, O‐donor ligands, we have looked into the electronic structure of N,O‐hybrid 1,10‐phenanthroline derived phosphine oxide ligands with various cyclic substitutions. The complexation behavior of these ligands with Am(III) and Eu(III) is evaluated using density functional theory calculations. Electronic structures divulge that in all these tetra‐coordinated complexes, the oxygen atoms bind strongly than the nitrogen atoms. The substitution of a benzene ring on the phosphorus increases the selectivity whereas the binding is enhanced by the substitution of cyclopentane. QTAIM analysis portrays the weak C−H−O interactions present in these complexes. Energy decomposition analysis in gas and solvent phases estimates the weak stabilizing interactions in these complexes. All the calculations bring out the salient structural features of these ligands which will further enrich our understanding of the tailoring of ligands for efficient uptake of Am (III) ions over Eu (III) ions.

Biomineralized Materials as Model Systems for Structural Composites: Intracrystalline Structural Features and Their Strengthening and Toughening Mechanisms.

Biomineralized composites, which are usually composed of microscopic mineral building blocks organized in 3D intercrystalline organic matrices, have evolved unique structural designs to fulfill mechanical and other biological functionalities. While it has been well recognized that the intricate architectural designs of biomineralized composites contribute to their remarkable mechanical performance, the structural features within and corresponding mechanical properties of individual mineral building blocks are often less appreciated in the context of bio‐inspired structural composites. The mineral building blocks in biomineralized composites exhibit a variety of salient intracrystalline structural features, such as, organic inclusions, inorganic impurities (or trace elements), crystalline features (e.g., amorphous phases, single crystals, splitting crystals, polycrystals, and nanograins), residual stress/strain, and twinning, which significantly modify the mechanical properties of biogenic minerals. In this review, recent progress in elucidating the intracrystalline structural features of three most common biomineral systems (calcite, aragonite, and hydroxyapatite) and their corresponding mechanical significance are discussed. Future research directions and corresponding challenges are proposed and discussed, such as the advanced structural characterizations and formation mechanisms of intracrystalline structures in biominerals, amorphous biominerals, and bio‐inspired synthesis.

Fabrication of CoWO4/PANI composite as electrode material for energy storage applications

The fabrication of novel electrochemical devices with salient structural features, morphology and porosity is necessary to enhance their efficiency and shelf life. The construction of electrodes using binary metal oxides (BMOs) with conducting polymers (CPs) is noted to be one of the promising strategies to enhance the electrochemical performance of supercapacitors (SCs). Herein, we reported an in-situ chemical oxidative polymerization synthetic route to prepare CoWO4/PANI composite. The structure, crystalline phase, surface morphology and electrochemical (EC) attributes of the as-synthesized material were studied by different spectral and analytical techniques. The cloud-like CoWO4/PANI electrode material exhibited a maximum specific capacitance (Csp) of 653 Fg-1 at the scan rate of 5 mVs−1. The prominent cycling stability of CoWO4/PANI with 93.3% retention even after 5000 cycles at the current density of 1 Ag-1 was also observed. The study revealed cost-effective synthesis and fabrication of high-performance CoWO4/PANI electrode material for energy storage devices.

Kinship in Contemporary Japanese Society : An Overview

The paper seeks to analyse the salient features of the kinship structure in Japan and how it plays it is significant role to form the traditional Japanese values. The paper also tries to analyse the changing nature of the kinship system in contemporary Japanese society from its traditional family (IE) system. The paper follows the qualitative method of research where the data has been collected from both academic and non-academic sources. By analysing the kinship structure of different periods in Japan, the paper finds that during the IE family system of the Tokugawa period, Japan gives less weight to kinship relations than other Asian countries. The feeling on son, either related to blood or adopted, marks the major distinction with other societies to find out the kin and non-kin. Following the IE system, the paper finds another two major events behind the weaken kinship structure in Japan: the emergence of koseki since the Meiji restoration and the rise of corporate culture during the contemporary period.
 Social Science Review, Vol. 37(2), Dec 2020 Page 105-124

Protein conformational switch discerned via network centrality properties

Network analysis has emerged as a powerful tool for examining structural biology systems. The spatial organization of the components of a biomolecular structure has been rendered as a graph representation and analyses have been performed to deduce the biophysical and mechanistic properties of these components. For proteins, the analysis of protein structure networks (PSNs), especially via network centrality measurements and cluster coefficients, has led to identifying amino acid residues that play key functional roles and classifying amino acid residues in general. Whether these network properties examined in various studies are sensitive to subtle (yet biologically significant) conformational changes remained to be addressed. Here, we focused on four types of network centrality properties (betweenness, closeness, degree, and eigenvector centralities) for conformational changes upon ligand binding of a sensor protein (constitutive androstane receptor) and an allosteric enzyme (ribonucleotide reductase). We found that eigenvector centrality is sensitive and can distinguish salient structural features between protein conformational states while other centrality measures, especially closeness centrality, are less sensitive and rather generic with respect to the structural specificity. We also demonstrated that an ensemble-informed, modified PSN with static edges removed (which we term PSN*) has enhanced sensitivity at discerning structural changes.

Multi-scale architecture of archaeal chromosomes

Chromosome conformation capture (3C) technologies have identified topologically associating domains (TADs) and larger A/B compartments as two salient structural features of eukaryotic chromosomes. These structures are sculpted by the combined actions of transcription and structural maintenance of chromosomes (SMC) superfamily proteins. Bacterial chromosomes fold into TAD-like chromosomal interaction domains (CIDs) but do not display A/B compartment-type organization. We reveal that chromosomes of Sulfolobus archaea are organized into CID-like topological domains in addition to previously described larger A/B compartment-type structures. We uncover local rules governing the identity of the topological domains and their boundaries. We also identify long-range loop structures and provide evidence of a hub-like structure that colocalizes genes involved in ribosome biogenesis. In addition to providing high-resolution descriptions of archaeal chromosome architectures, our data provide evidence of multiple modes of organization in prokaryotic chromosomes and yield insights into the evolution of eukaryotic chromosome conformation.

Seasonal changes in reproductive anatomy and gonadal hormone concentrations of African penguins (Spheniscus demersus)

Several standard descriptions of the avian male and female reproductive tract have been reported, including effects of age, stage of reproductive maturity and gonadal hormone concentrations. Limited information on penguin reproductive biology and a lack of information on the African penguin (Spheniscus demersus) necessitated a detailed description of salient structural features of this species and provided an opportunity to evaluate seasonal changes in gonadal steroid hormone concentrations. Tissues from 36 males (adults and juveniles) and 29 females (adults and juveniles) were used for macro-anatomical descriptions and histology of the testes and ovaries. In addition, concentrations of gonadal steroid hormones for eight captive African penguins (four females and four males) were quantified during two breeding and one non-breeding season. The testes were asymmetrical, with the right testis having smaller dimensions compared to the left testis. Marked spermatogenic cellular associations and spermatid developmental stages were present in adult testes only during the breeding season. There was variation in the dimensions of the single ovary during follicular development related to the age and breeding status of the females. Testosterone, dihydrotestosterone, and estradiol concentrations fluctuated during the breeding and non-breeding periods, with males and females having similar steroid concentrations. The results from this study confirm that the breeding status in African penguins can be deduced based on testicular and ovarian histological structures. The results from the present study focused on African penguin reproductive biology should be considered in management strategies for the conservation of the species.

WebNet: A biomateriomic three-dimensional spider web neural net

Spiders, silks and webs are abundant in most ecosystems, suggesting that they are a significant evolutionary success. The structures they build – in particular various forms of webs – have intrigued engineers for a long time, and elucidated inspiration for new mechanical designs of de novo materials. Here we report the development of a biomateriomic neural network based constitutive model to describe the mechanical features, such as strength and toughness, of a 3D spider web depending on salient structural features, specifically: average fiber lengths, fiber orientations, web connectivity, and web density. In particular, we focus our study on the structural and mechanical properties of a Cyrtophora citricola spider web, and report a method to derive the neural net model directly from the experimental–computational mesoscale modeling, using a novel data augmentation method. Our machine learning model captures the complex biomateriomical mechanics of spider webs. More generally, approaches such as reported here can be useful to describe the intricate relationships in other hierarchical materials, and provide a basis to develop multiscale models, and bridge experimental data with computational and theoretical modeling efforts.

Structure features of peptide-type SARS-CoV main protease inhibitors: Quantitative structure activity relationship study

In the present work, an extensive QSAR (Quantitative Structure Activity Relationships) analysis of a series of peptide-type SARS-CoV main protease (MPro) inhibitors following the OECD guidelines has been accomplished. The analysis was aimed to identify salient and concealed structural features that govern the MPro inhibitory activity of peptide-type compounds. The QSAR analysis is based on a dataset of sixty-two peptide-type compounds which resulted in the generation of statistically robust and highly predictive multiple models. All the developed models were validated extensively and satisfy the threshold values for many statistical parameters (for e.g. R2 ​= ​0.80–0.82, Q2loo ​= ​0.74–0.77, Q2LMO ​= ​0.66–0.67). The developed QSAR models identified number of sp2 hybridized Oxygen atoms within seven bonds from aromatic Carbon atoms, the presence of Carbon and Nitrogen atoms at a topological distance of 3 and other interrelations of atom pairs as important pharmacophoric features. Hence, the present QSAR models have a good balance of Qualitative (Descriptive QSARs) and Quantitative (Predictive QSARs) approaches, therefore useful for future modifications of peptide-type compounds for anti- SARS-CoV activity.

Observation of flat bands in twisted bilayer graphene

Transport experiments in twisted bilayer graphene revealed multiple superconducting domes separated by correlated insulating states. These properties are generally associated with strongly correlated states in a flat mini-band of the hexagonal moir\'e superlattice as it was predicted by band structure calculations. Evidence for such a flat band comes from local tunneling spectroscopy and electronic compressibility measurements, reporting two or more sharp peaks in the density of states that may be associated with closely spaced van Hove singularities. Direct momentum resolved measurements proved difficult though. Here, we combine different imaging techniques and angle resolved photoemission with simultaneous real and momentum space resolution (nano-ARPES) to directly map the band dispersion in twisted bilayer graphene devices near charge neutrality. Our experiments reveal large areas with homogeneous twist angle that support a flat band with spectral weight that is highly localized in momentum space. The flat band is separated from the dispersive Dirac bands which show multiple moir\'e hybridization gaps. These data establish the salient features of the twisted bilayer graphene band structure.

Voxelized Atomic Structure Potentials: Predicting Atomic Forces with the Accuracy of Quantum Mechanics Using Convolutional Neural Networks.

This paper introduces voxelized atomic structure (VASt) potentials as a machine learning (ML) framework for developing interatomic potentials. The VASt framework utilizes a voxelized representation of the atomic structure directly as the input to a convolutional neural network (CNN). This allows for high-fidelity representations of highly complex and diverse spatial arrangements of the atomic environments of interest. The CNN implicitly establishes the low-dimensional features needed to correlate each atomic neighborhood to its net atomic force. The selection of the salient features of the atomic structure (i.e., feature engineering) in the VASt framework is implicit, comprehensive, automated, scalable, and highly efficient. The calibrated convolutional layers learn the complex spatial relationships and multibody interactions that govern the physics of atomic systems with remarkable fidelity. We show that VASt potentials predict highly accurate forces on two phases of silicon carbide and the thermal conductivity of silicon over a range of isotropic strain.

Visual privacy-preserving level evaluation for multilayer compressed sensing model using contrast and salient structural features

Recognition and classification tasks in images or videos are ubiquitous, but they can lead to privacy issues. People increasingly hope that camera systems can record and recognize important events and objects, such as real-time recording of traffic conditions and accident scenes, elderly fall detection, and in-home monitoring. However, people also want to ensure these activities do not violate the privacy of users or others. The sparse representation classification and recognition algorithms based on compressed sensing (CS) are robust at recognizing human faces from frontal views with varying expressions and illuminations, as well as occlusions and disguises. This is a potential way to perform recognition tasks while preserving visual privacy. In this paper, an improved Gaussian random measurement matrix is adopted in the proposed multilayer CS (MCS) model to realize multiple image CS and achieve a balance between visual privacy-preserving and recognition tasks. The visual privacy-preserving level evaluation for MCS images has important guiding significance for image processing and recognition. Therefore, we propose an image visual privacy-preserving level evaluation method for the MCS model (MCS-VPLE) based on contrast and salient structural features. The basic concept is to use the contrast measurement model based on the statistical mean of the asymmetric alpha-trimmed filter and the salient generalized center-symmetric local binary pattern operator to extract contrast and salient structural features, respectively. The features are fed into a support vector regression to obtain the image quality score, and the fuzzy c-means algorithm is used for clustering to obtain the final evaluated image visual privacy-preserving score. Experiments on three constructed databases show that the proposed method has better prediction effectiveness and performance than conventional methods.