Anhui University

Biomedical signals, encapsulating vital physiological information, are pivotal in elucidating human traits and conditions, serving as a cornerstone for advancing human-machine interfaces. Nonetheless, the fidelity of biomedical signal interpretation is frequently compromised by pervasive noise sources such as skin, motion, and equipment interference, posing formidable challenges to precision recognition tasks. Concurrently, the burgeoning adoption of intelligent wearable devices illuminates a societal shift towards enhancing life and work through technological integration. This surge in popularity underscores the imperative for efficient, noise-resilient biomedical signal recognition methodologies, a quest that is both challenging and profoundly impactful. This study proposes a novel approach to enhancing biomedical signal recognition. The proposed approach employs a hierarchical information bottleneck mechanism within SNNs, quantifying the mutual information in different orders based on the depth of information flow in the network. Subsequently, these mutual information, together with the network’s output and category labels, are restructured based on information theory principles to form the loss function used for training. A series of theoretical analyses and substantial experimental results have shown that this method can effectively compress noise in the data, and on the premise of low computational cost, it can also significantly outperform its vanilla counterpart in terms of classification performance.

Magnetic antiskyrmions, the anti‐quasiparticles of magnetic skyrmions, possess alternating Bloch‐ and Néel‐type spin spirals, rendering them promising for advanced spintronics‐based information storage. To date, antiskyrmions are demonstrated in a few bulk materials featuring anisotropic Dzyaloshinskii–Moriya interactions and a limited number of artificial multilayers. Identifying novel film materials capable of hosting isolated antiskyrmions is critical for memory applications in topological spintronics. Herein, the formation of room‐temperature antiskyrmions in single ferrimagnetic CoHo rare‐metal alloy films of varying thicknesses, observed using Lorentz transmission electron microscopy is reported. Furthermore, rotating magnetic fields (H) are proposed to facilitate antiskyrmion nucleation and enhance their areal density by an order of magnitude compared to that in the same area under individual vertical H. In addition, experimental and phenomenological analysis confirm that antiskyrmion nucleation can be attributed to spin reorientation involving spontaneous canted magnetism, as evidenced by polarized neutron reflectometry. Micromagnetic simulations further show that the antiskyrmion density significantly depends on the magnitude of the rotating field. These findings expand the family of known antiskyrmion‐hosting materials and provide insights into their formation mechanisms, thus serving as a basis for their application in topological spintronics.

Understanding the phase structure‐dependent catalytic performance is of great significance for the investigation of advanced electrocatalysts. At present, research in phase engineering of metal materials for electrocatalysis predominantly concentrates on the iron group, platinum group, and coinage group metals with A1‐, A2‐, and A3‐ type structures. However, the investigation of metal phase engineering beyond the above metal group with other types of structures for electrocatalysis is still poorly explored. Herein, using tungsten as a substrate to support iridium, it is shown that iridium‐embedded tungsten with diverse crystal phase structure (referred to as Ir/α‐W for A2‐type structure and Ir/β‐W for A‐15 type structure) exhibits distinct catalytic activity for hydrogen oxidation reactions (HOR) in alkaline medium. Notably, the mass‐normalized exchange current density (j0, m) of noble metal iridium in Ir/α‐W (518.3 A g⁻¹Ir) is ≈1.8 times and 16.4 times higher than that of Ir/β‐W and Ir/C, respectively. In‐depth mechanistic studies suggest that the enhanced HOR performance on Ir/α‐W is attributed to the enhanced connectivity of the H‐bond network as well as the synergistic optimization of the adsorption binding energies of H and OH intermediate species. This study can inspire more scientific interest on the exploration of metal phase engineering for electrocatalysis.

Multiphoton excited fluorescence (MPEF) imaging has emerged as a powerful tool for visualizing biological processes with high spatial and temporal resolution. Metal‐organic frameworks (MOFs), a class of porous materials composed of metal ions or clusters coordinated with organic ligands, have recently gained attention for their unique optical properties and potential applications in MPEF imaging. This review provides a comprehensive overview of the design, synthesis, and applications of multiphoton excited fluorescence imaging using MOFs. We discuss the principles behind the fluorescence behavior of MOFs, explore strategies to enhance their photophysical properties, and showcase their applications in bioimaging. Additionally, we address the current challenges and future prospects in this rapidly evolving field, highlighting the potential of multiphoton excited fluorescence imaging by MOFs for advancing our understanding of complex biological processes.

Pathogenic bioaerosols are critical for outbreaks of airborne disease; however, rapidly and accurately identifying pathogens directly from complex air environments remains highly challenging. We present an advanced method that combines open-set deep learning (OSDL) with single-cell Raman spectroscopy to identify pathogens in real-world air containing diverse unknown indigenous bacteria that cannot be fully included in training sets. To test and further enhance identification, we constructed the Raman datasets of aerosolized bacteria. Through optimizing OSDL algorithms and training strategies, Raman-OSDL achieves 93% accuracy for five target airborne pathogens, 84% accuracy for untrained air bacteria, and 36% reduction in false positive rates compared to conventional close-set algorithms. It offers a high detection sensitivity down to 1:1000. When applied to real air containing >4600 bacterial species, our method accurately identifies single or multiple pathogens simultaneously within an hour. This single-cell tool advances rapidly surveilling pathogens in complex environments to prevent infection transmission.

Emberiza buntings (Aves: Emberizidae) exhibit extensive diversity and rapid diversification within the Old World, particularly in the eastern Palearctic, making them valuable models for studying rapid radiation among sympatric species. Despite their ecological and morphological diversity, there remains a significant gap in understanding the genomic underpinnings driving their rapid speciation. To fill this gap, we assembled high‐quality chromosome‐level genomes of five representative Emberiza species ( E. aureola , E. pusilla , E. rustica , E. rutila and E. spodocephala ). Comparative genomic analysis revealed distinct migration‐related evolutionary adaptations in their genomes, including variations in lipid metabolism, oxidative stress response, locomotor ability and circadian regulation. These changes may facilitate the rapid occupation of emerging ecological niches and provide opportunities for species diversification. Additionally, these five species exhibited abnormal abundances of long terminal repeat retrotransposons (LTRs), comprising over 20% of their genomes, with insertion times corresponding to their divergence (~2.5 million years ago). The presence of LTRs influenced genome size, chromosomal structure and single‐gene expression, suggesting their role in promoting the rapid diversification of Emberiza species. These findings offer valuable insights into the adaptive radiation of Emberiza and establish a robust theoretical foundation for further exploration of the patterns and mechanisms underlying their diversification.

Mimicking the superstructures and properties of spherical biological encapsulants such as viral capsids¹ and ferritin² offers viable pathways to understand their chiral assemblies and functional roles in living systems. However, stereospecific assembly of artificial polyhedra with mechanical properties and guest-binding attributes akin to biological encapsulants remains a formidable challenge. Here we report the stereospecific assembly of dynamic supramolecular snub cubes from 12 helical macrocycles, which are held together by 144 weak C–H hydrogen bonds³. The enantiomerically pure snub cubes, which have external diameters of 5.1 nm, contain 2,712 atoms and chiral cavities with volumes of 6,215 ų. The stereospecific assembly of left- and right-handed snub cubes was achieved by means of a hierarchical chirality transfer protocol⁴, which was streamlined by diastereoselective crystallization. In addition to their reversible photochromic behaviour, the snub cubes exhibit photocontrollable elasticity and hardness in their crystalline states. The snub cubes can accommodate numerous small guest molecules simultaneously and encapsulate two different guest molecules separately inside the uniquely distinct compartments in their frameworks. This research expands the scope of artificial supramolecular assemblies to imitate the chiral superstructures, dynamic features and binding properties of spherical biomacromolecules and also establishes a protocol for construction of crystalline materials with photocontrollable mechanical properties.

This work provides an alternative confirmation of Dirac localization in the photonic crystal fiber proposed in Light Sci. Appl. 4, e304 (2015)LSAIAZ2047-753810.1038/lsa.2015.77. In weakly guiding fibers, Maxwell equations can be reduced to the wave equation, which can then be further reduced to the Helmholtz equation for time-harmonic waves. The Dirac point in the band structure of the Helmholtz equation in periodic media is analyzed. Wave localization at the Dirac point is numerically demonstrated, which confirms the finding of Light Sci. Appl. 4, e304 (2015)LSAIAZ2047-753810.1038/lsa.2015.77. Coupling of Dirac modes is numerically simulated, and the results show remote coupling is feasible. Coupling strength is shown to decrease in an algebraic form with distance. This feature of coupling can be attributed to the massless property of Dirac quasiparticles, which is a precise analog of Klein tunneling of massless relativistic electrons in graphene.

Antivortices have potential applications in future nano-functional devices, yet the formation of isolated antivortices traditionally requires nanoscale dimensions and near-zero magnetocrystalline anisotropy, limiting their broader application. Here, we propose an approach to forming antivortices in multiferroic ε-Fe2O3 with the coalescence of misaligned grains. By leveraging misaligned crystal domains, the large magnetocrystalline anisotropy energy is counterbalanced, thereby stabilizing the ground state of the antivortex. This method overcomes the traditional difficulty of observing isolated antivortices in micron-sized samples. Stable isolated antivortices were observed in truncated triangular multiferroic ε-Fe2O3 polycrystals ranging from 2.9 to 16.7 µm. Furthermore, the unpredictability of the polarity of the core was utilized as a source of entropy for designing physically unclonable functions. Our findings expand the range of antivortex materials into the multiferroic perovskite oxides and provide a potential opportunity for ferroelectric polarization control of antivortices.

Controlling symmetrical or asymmetrical growth has allowed a series of novel nanomaterials with prominent physicochemical properties to be produced. However, precise and continuous size growth based on a preserved template has long been a challenging pursuit, yet little has been achieved in terms of manipulation at the atomic level. Here, a correlated silver cluster series has been established, enabling atomically precise manipulation of symmetrical and asymmetrical surface structure expansions of metal nanoclusters. Specifically, the C3-axisymmetric Ag29(BDTA)12(PPh3)4 nanocluster underwent symmetrical and asymmetrical surface structure expansions via an acid-mediated synthetic procedure, giving rise to C3-axisymmetric Ag32(BDTA)12(PPh3)10 and C1-axisymmetric Ag33(BDTA)12(PPh3)11, respectively. In addition, structural transformations, including structural degradation from Ag32 to Ag29 and asymmetrical structural expansion from Ag32 to Ag33, were rationalized theoretically. More importantly, the asymmetrically structured Ag33 nanoclusters followed a chiral crystallization mode, and their crystals displayed high optical activity, derived from CD and CPL characterization. This work not only provides an important model for unlocking the symmetrical/asymmetrical size growth mechanism at the atomic level but also pioneers a promising approach to activate the optical activity of cluster-based nanomaterials.

Artificial neurons with leaky rectified linear unit (L‐ReLU) function can effectively process negative information, enhancing the neuromorphic systems capbility to handle negative values. Memristive devices show great potential in building compact and bio‐plausible artificial neurons, however, a neuron device that supports L‐ReLU functions is still lacking. In this work, a compact L‐ReLU neuron is proposed based on a bipolar asymmetrical diffusive memristor. Utilizing intercalation and leveraging the migration diffusion ratio of Ag ions, devices meeting the characteristics of the L‐ReLU function are fabricated (Ag/TaOx/SiOx/Pt). Thus, the constructed neuron can encode the positive and negative input information into positive and negative spikes, respectively, with L‐ReLU‐like response function. In addition, to address the problem that the frequency only has positive values, a vector frequency (VF) is defined to describe the frequency of the positive and negative spikes. Moreover, a convolutional neural network and You‐Only‐Look‐Once version 3 (YOLOv3) network are constructed with the L‐ReLU neurons for gesture language recognition and target detection tasks, respectively. The network based on L‐ReLU neurons can avoid negative information squandering, showing a higher inference accuracy than the network with ReLU neurons. The results show that the neurons can offer a promising platform for building high‐accuracy neuromorphic systems.

Background Echinococcosis, a parasitic disease caused by the larvae of the Echinococcus parasite, poses a serious threat to human health. Medication is an indispensable means of treatment for Echinococcosis; however, due to the less satisfactory efficacy of single drugs, identifying effective drug combinations for the treatment of Echinococcosis is essential. Yet, current predictive models for drug synergy in Echinococcosis face accuracy challenges due to data scarcity, method limitations, and insufficient feature representation. Objective This work aims to design an end-to-end method to predict drug synergistic combinations, which enables efficient and accurate identification of drug combinations against Echinococcosis. Methods In this work, an end-to-end method, named DSPE, is proposed for predicting anti- Echinococcosis drug synergistic combinations. In DSPE, a dataset of Echinococcosis drug synergistic combinations is constructed by retrieving and extracting information from related scientific articles. Further, DSPE employs a residual graph attention network to deeply analyze drug characteristics and their interactions, thereby enhancing the performance of deep learning models. It also explores the protein-protein interaction network related to Echinococcosis, using node2vec combined with an attention mechanism to efficiently encode disease features. Finally, it predicts the synergy of drug combinations based on the Bliss score by integrating drug combinations and disease features. Results Experimental evidence shows that DSPE outperforms five state-of-the-art algorithms in predicting drug combination effects by leveraging disease-target information and single-agents for the treatment. Conclusion DSPE improves prediction accuracy and addresses the issue of data scarcity for new diseases, offering new insights and methods for the development of treatment plans for parasitic diseases in the future.

We investigate the performance of an active reconfigurable intelligent surface (RIS)-assisted mixed radio frequency (RF)-terahertz (THz) relaying system, where the RF signal reaches the relay through the active RIS and is then transmitted to the user via the THz channel. Under this scenario, we analyze the system performance with the relay employing amplify-and-forward (AF) and decode-and-forward (DF) protocols. More specifically, we derive the exact expressions for the cumulative distribution function (CDF) of the end-to-end signal-to-noise ratio (SNR) for both relaying protocols. Based on this, we obtain the exact expressions for the outage probability, average bit error rate (ABER), and average channel capacity (ACC). Furthermore, to gain deeper insights, we derive the asymptotic expressions at high SNRs and obtain diversity order (DO) of the system. Moreover, we extend the analysis to the variable gain relaying scheme. The findings reveal that the DOs for both relaying protocols are determined by the THz channel parameters, and the DO under the AF relaying protocol is twice that of the DF relaying protocol. Finally, we validate that active RIS (A-RIS) can more effectively assist the performance of the mixed RF-THz relaying system compared to passive RIS.

The inherent trade‐off between permeability and selectivity has constrained further improvement of passive linear force‐electric conversion performance in nanofluidic pressure sensors. To overcome this limitation, a 3D nanofluidic membrane with high mechanical strength utilizing aramid nanofibers/carbon nanofiber (ANF/CNF) dual crosslinking is developed. Due to the abundant surface functional groups of CNF and the high mechanical strength of ANF, this large‐scale integrated 3D nanofluidic membrane exhibits advantages of high flux, high porosity, and short ion transport path, demonstrating superior force‐electric response compared to conventional 1D and 2D configurations. The enhancement mechanism of the ANF/CNF membrane is systematically investigated through experimental results and theoretical calculations. The optimized device has a sensitivity of 111 nA cm⁻² kPa⁻¹, a response/recovery time of 63/68 ms, and a stability of 45 000 cycles. This study successfully overcomes the inherent performance limitations of traditional nanofluidic membranes, offering promising potential for applications across artificial intelligence, the Internet of Things, and smart wearable devices.

Ternary isodual codes and their duals are shown to support 3-designs under mild symmetry conditions. These designs are held invariant by a double cover of the permutation part of the automorphism group of the code. Examples of interest include extended quadratic residues (QR) codes of lengths 14 and 38 whose automorphism groups are PSL(2, 13) and PSL(2, 37), respectively. We also consider Generalized Quadratic Residue (GQR) codes in the sense of Lint and MacWiliams (IEEE Trans Inf Theory 24(6): 730-737,1978). These codes are the abelian generalizations of the Quadratic Residue (QR) codes which are cyclic. We construct them as row span of a Jacobsthal matrix. In lengths 50 and 26 we obtain 3-designs invariant under a double cover of $P{\Sigma }L(2,49),$ and $P{\Sigma }L(2,25),$ respectively. In addition, from block orbits of these 3-designs we construct a number of other 3-designs and 2-designs. Finally, we apply the same construction to the binary extended GQR code of length 82.

Synonymous mutations, once considered neutral, are now understood to have significant implications for a variety of diseases, particularly cancer. It is indispensable to identify these driver synonymous mutations in human cancers, yet current methods are constrained by data limitations. In this study, we initially investigate the impact of sequence-based features, including DNA shape, physicochemical properties and one-hot encoding of nucleotides, and deep learning-derived features from pre-trained chemical molecule language models based on BERT. Subsequently, we propose EPEL, an effect predictor for synonymous mutations employing ensemble learning. EPEL combines five tree-based models and optimizes feature selection to enhance predictive accuracy. Notably, the incorporation of DNA shape features and deep learning-derived features from chemical molecule represents a pioneering effect in assessing the impact of synonymous mutations in cancer. Compared to existing state-of-the-art methods, EPEL demonstrates superior performance on independent test datasets. Furthermore, our analysis reveals a significant correlation between effect scores and patient outcomes across various cancer types. Interestingly, while deep learning methods have shown promise in other fields, their DNA sequence representations do not significantly enhance the identification of driver synonymous mutations in this study. Overall, we anticipate that EPEL will facilitate researchers to more precisely target driver synonymous mutations. EPEL is designed with flexibility, allowing users to retrain the prediction model and generate effect scores for synonymous mutations in human cancers. A user-friendly web server for EPEL is available at http://ahmu.EPEL.bio/ .

A wideband spatiotemporally modulated nonmagnetic circulator composed of three time‐varying resonators in Y‐topology is presented in this paper. The triple resonators are designed with coupled microstrip lines and connected to a central node. By applying a time‐modulated signal with a 120° progressive phase shift to the resonators, the angular momentum bias is applied to the structure, thereby achieving the circulator response. To enhance the bandwidth, a step impedance resonator is added to the end of each coupled microstrip line resonator, and therefore achieves broad isolation band for the circulator. As validation, a wideband nonmagnetic circulator with a center frequency of 1.0 GHz was designed, fabricated and measured. The measurement results agree well with the simulations, which the 20‐dB isolation bandwidth is 53 MHz, while the maximum in‐band isolation can achieve 34 dB, and the insertion loss is 1.9 dB.