Search

Although the precise symmetric forward and feedback connections between neurons are thought to be impossible in the brain, most existing deep spiking neural networks (SNNs) spatiotemporal learning algorithms still require such a strong architectural constraint to complete tasks. Besides that, as an expected feature for specialized neuromorphic hardware to be deployed, effective online learning of deep SNNs for training spatiotemporal data is still lacking. To address these issues, an effective biologically plausible asymmetric spatiotemporal online learning algorithm called ASTOL is proposed for training deep SNNs in real time. ASTOL could learn the spatial and temporal features simultaneously in real time, without the precise weights symmetric backpropagation and the need to know the duration of spatiotemporal data in advance. Experimental results show that the proposed ASTOL algorithm achieves comparable performance in real-time learning on the rate coding MNIST data set and the temporal coding music MedleyDB data set and speech TIDIGITS data set with other state-of-the-art algorithms with a simpler learning mechanism way as it avoids transport of synaptic weights. Besides that, the experiment on MNIST data sets also shows that ASTOL can use simpler network structures to achieve good performance faster.

Inorganic phosphate (Pi) is a fundamental macronutrient for all living organisms, the homeostasis of which is critical for numerous biological activities1–3. As the only known human Pi exporter to date, XPR1 has an indispensable role in cellular Pi homeostasis4,5. Dysfunction of XPR1 is associated with neurodegenerative disease6–8. However, the mechanisms underpinning XPR1-mediated Pi efflux and regulation by the intracellular inositol polyphosphate (InsPP) sensor SPX domain remain poorly understood. Here we present cryo-electron microscopy structures of human XPR1 in Pi-bound closed, open and InsP6-bound forms, revealing the structural basis for XPR1 gating and regulation by InsPPs. XPR1 consists of an N-terminal SPX domain, a dimer-formation core domain and a Pi transport domain. Within the transport domain, three basic clusters are responsible for Pi binding and transport, and a conserved W573 acts as a molecular switch for gating. In addition, the SPX domain binds to InsP6and facilitates Pi efflux by liberating the C-terminal loop that limits Pi entry. This study provides a conceptual framework for the mechanistic understanding of Pi homeostasis by XPR1 homologues in fungi, plants and animals.

A zinc-organic hybrid (1) with multifunctional room temperature phosphorescence (RTP) was synthesized. 1presents light/force-sensitive RTP properties due to the photochromic behavior from gray to light yellow and the transition from crystalline to amorphous state, respectively. Furthermore, inkless printing and information encryption models were successfully constructed to prove their widespread application prospect.

Sc2CO2MXene has potential applications in energy storage and optoelectronics due to its superior structure and excellent properties. The electronic properties, quantum capacitance, and photocatalytic activity of Cr-doped Sc2CO2under strain are studied by the density functional theory. Cr doping makes the system produce magnetism. The spin-down states of Sc2CO2–Cr under strain are direct semiconductors, while their spin-up states are indirect semiconductors. Sc2CO2–Cr under +5, −5, −3, and −2% strains in an aqueous system are suitable for cathode material. A large voltage drastically modulates the type of electrode materials. Sc2CO2–Cr under strains from 0 to +2% can perform the oxygen evolution reaction at an alkaline environment, while the Sc2CO2–Cr system under strain is a good for CO2photocatalysis at pH 0 and 7.

The conversion of DNA 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) by TET enzymes represents a significant epigenetic modification, yet its role in early human embryos remains largely unknown. Here we showed that the early human embryo inherited a significant amount of 5hmCs from an oocyte, which unexpectedly underwent de novo hydroxymethylation during its growth. Furthermore, the generation of 5hmC in the paternal genome after fertilization roughly followed the maternal pattern, which was linked to DNA methylation dynamics and regions of sustained methylation. The 5hmCs persisted until the eight-cell stage and exhibited high enrichment at OTX2 binding sites, whereas knockdown of OTX2 in human embryos compromised the expression of early lineage genes. Specifically, the depletion of 5hmC affected the activation of embryonic genes, which was further evaluated by ectopically expressing mouse Tet3 in human early embryos. These findings revealed distinct dynamics of 5hmC and unravelled its multifaceted functions in early human embryonic development.

Friction remains as the primary mode of energy dissipation and components wear, and achieving superlubricity shows high promise in energy conservation and lifetime wear protection. The results in this work demonstrate that direct superlubricity combined with superlow wear can be realized for steel/Si3N4contacts on engineering scale when polyhydroxy alcohol solution was selectively modified by amino group. Macroscopic direct superlubricity occurs because 3-amino-1,2-propanediol molecules at the friction interface could be induced to rotate and adsorb vertically on the friction surface, forming in-situthick and dense molecular films to passivate the asperity contacts. Furthermore, amino modification is also conducive to improving the lubrication state from boundary to mixed lubrication regime by strengthening the intermolecular hydrogen bonding interaction, presenting enhanced load-bearing capability and reduced direct solid asperity contacts. Thus, direct superlow average friction of 0.01 combined with superlow wear are achieved simultaneously. The design principle of direct superlubricity and superlow wear in this work indeed offers an effective strategy to fundamentally improve energy efficiency and provide lifetime wear protection for moving mechanical assemblies.

While supervised learning of over-parameterized neural networks achieved state-of-the-art performance in image classification, it tends to over-fit the labeled training samples to give inferior generalization ability. Output regularization deals with over-fitting by using soft targets as additional training signals. Although clustering is one of the most fundamental data analysis tools for discovering general-purpose and data-driven structures, it has been ignored in existing output regularization approaches. In this article, we leverage this underlying structural information by proposing Cluster-based soft targets for Output Regularization (CluOReg). This approach provides a unified way for simultaneous clustering in embedding space and neural classifier training with cluster-based soft targets via output regularization. By explicitly calculating a class relationship matrix in the cluster space, we obtain classwise soft targets shared by all samples in each class. Results of image classification experiments under various settings on a number of benchmark datasets are provided. Without resorting to external models or designed data augmentation, we get consistent and significant reductions in classification error compared with other approaches, demonstrating that cluster-based soft targets effectively complement the ground-truth label.

The temporal credit assignment (TCA) problem, which aims to detect predictive features hidden in distracting background streams, remains a core challenge in biological and machine learning. Aggregate-label (AL) learning is proposed by researchers to resolve this problem by matching spikes with delayed feedback. However, the existing AL learning algorithms only consider the information of a single timestep, which is inconsistent with the real situation. Meanwhile, there is no quantitative evaluation method for TCA problems. To address these limitations, we propose a novel attention-based TCA (ATCA) algorithm and a minimum editing distance (MED)-based quantitative evaluation method. Specifically, we define a loss function based on the attention mechanism to deal with the information contained within the spike clusters and use MED to evaluate the similarity between the spike train and the target clue flow. Experimental results on musical instrument recognition (MedleyDB), speech recognition (TIDIGITS), and gesture recognition (DVS128-Gesture) show that the ATCA algorithm can reach the state-of-the-art (SOTA) level compared with other AL learning algorithms.

Although organic–inorganic hybrid Mn2+halides have advanced significantly, achieving high stability and narrow-band emission remains enormously challenging owing to the weak ionic nature and soft crystal lattice of the halide structure. To address these issues, we proposed a cationic engineering strategy of long-range cation π···π stacking and C–H···π interactions to simultaneously improve the crystal structural stability and rigidity. Herein, two organic zero-dimensional (0D) manganese halide hybrids of (BACQ)2MnX4[BACQ = 4-(butylamino)-7-chloroquinolin-1-ium; X = Cl and Br] were synthesized. (BACQ)2MnX4display strong green-light emissions with the narrowest full width at half-maximum (fwhm) of 39 nm, which is significantly smaller than those of commercial green phosphor β-SiAlON:Eu2+and most of reported manganese halides. Detailed Hirshfeld surface analyses demonstrate the rigid environment around the [MnX4]2–units originating from the interactions between [BACQ]+. The rigid crystal structure weakens the electron–phonon coupling and renders narrow fwhm of these manganese halides, which is further confirmed by temperature-dependent emission spectra. Remarkably, (BACQ)2MnX4realizes outstanding structural and luminescence stabilities in various extreme environments. Benefiting from the excellent performance, these Mn2+halides are used to assemble light-emitting diodes with a wide color gamut of 105% of the National Television System Committee 1931 standard, showcasing the advanced applications in liquid-crystal-display backlighting.

In clinical practice, the treatment of colon cancer is faced with the dilemma of metastasis and recurrence, which is related to immunosuppression and hypoxia. Immune checkpoint blockade (ICB) is a negative regulatory pathway of immunity. Immune checkpoint blockade (ICB) is an important immunotherapy method. However, inadequate immunogenicity reduces the overall response rate of ICB. In this study, a tumor microenvironment-responsive nanomedicine (Cu-FACD@MnO2@FA) was prepared to increase host immune response and increase intracellular oxygen levels. Cu-FACD@MnO2@FA preferentially enriched at the tumor site, combined with the immune checkpoint inhibitor alpha PD-L1, induced sufficient immunogenicity to treat colon cancer. Immunofluorescence detection of tumor cells and tissues showed that the expression of hypoxa-inducing factor 1α was significantly down-regulated after treatment and the expression of immunoactivity-related proteins was significantly changed. In vivotreatment in a bilateral tumor mouse model showed complete ablation of the primary tumor and efficient inhibition of the distal tumor. In this study, for the first time, the oxygenation effects of MnO2-coated Cu-doped carbon dots and chemodynamic therapy and a strategy of combining with immuno-blocking therapy were used for treating colon cancer.

Herein, ultrathin CoOx/BiO2–xnanosheets (∼4 nm) have been in situ synthesized by using a microwave-assisted hydrothermal method. The amount-optimized CoOx/BiO2–xnanosheets display 4.7-fold higher photocatalytic activity than classical P25 TiO2for 2-chlorophenol (2-CP) degradation under full-spectrum irradiation with exceptional mineralization. The high photoactivity is attributed to the excitation-wavelength-dependent O2activation involved with the mainly produced •O2–via charge separation and 1O2through energy transfer with ultraviolet–visible (UV–vis, <450 nm) and visible-near-infrared (vis-NIR, >550 nm) light irradiation, respectively. Moreover, the modified CoOxwith high dispersion could effectively trap photogenerated holes, thereby promoting charge separation and also facilitating the formation of 1O2from •O2–by holes. By means of liquid chromatography tandem mass spectra and theoretical simulation, the generated nucleophilic •O2–and electrophilic 1O2radicals exhibit distinct preferences in attacking different sites of 2-CP, resulting in direct ring-opening processes and ultimately enabling rapid mineralization.

Oxide-dispersion-strengthened (ODS) reduced-activation ferritic/martensitic (RAFM) steels are considered as most promising structural materials for fusion reactors and other advanced nuclear systems due to its excellent irradiation resistance under high dose irradiations. However, the extreme complex synergistic effect between transmuted helium/hydrogen and displacement damage which most seriously affects the structure and mechanical properties of ODS-RAFM steels is very difficult to understand, especially the role of hydrogen still remains unclear. Taking advantage of very sensitive to small vacancy clusters of Positron annihilation spectroscopy (PAS) method, we conducted irradiation defect study on ODS-RAFM, RAFM and its corresponding model alloy Fe–9Cr and α-Fe to investigate their anti-swelling properties at the early stage of irradiation with hydrogen-helium synergism and the mechanism. Three distinct irradiation schemes including (1) single Fe ion beam, (2) simultaneous Fe and He (named Fe + He) ion beam, (3) subsequent H ion injection after Fe + He (named Fe + He/H) ion beam irradiation were performed. Hydrogen-helium synergistic effects on vacancy evolution within these steels were explored combining the first-principles calculation. The implantation of He was observed to significantly increase defect concentration among all four steels, while H exhibited interestingly complexity on affecting the evolution of helium bubbles. The H post-injection increased defect concentration obviously in the materials with abundant sinks, but improved poorly in ones with rare sinks. Through the systematic study of the interactions between H, He, and vacancies, it was revealed that even at the early irradiation stage, ODS-RAFM steels has already exhibited excellent irradiation resistance compared to other alloys across three irradiation schemes.

The metastasis and recurrence of cancer are related to immunosuppression and hypoxia in the tumor microenvironment. Activating immune activity and improving the hypoxic environment face essential challenges. This paper reports on a multifunctional nanomaterial, HSCCMBC, that induces immunogenic cell death through powerful photodynamic therapy/chemodynamic therapy synergistic antitumor effects. The tumor microenvironment changed from the immunosuppressive type to immune type, activated the immune activity of the system, decomposed hydrogen peroxide to generate oxygen based on Fenton-like reaction, and effectively increased the level of intracellular O2with the assistance of 3-bromopyruvate, a cell respiratory inhibitor. The structure and composition of HSCCMBC were characterized by transmission electron microscopy, X-ray photoelectron spectroscopy, X-ray diffraction, infrared spectroscopy, etc. Oxygen probe RDPP was used to investigate the oxygen level inside and outside the cell, and hydroxyl radical probe tetramethylbenzidine was used to investigate the Fenton-like reaction ability. The immunofluorescence method investigated the expression of various immune markers and hypoxia-inducing factors in vitro and in vivo after treatment. In vitro and in vivo experiments indicate that HSCCMBC is an excellent antitumor agent and is expected to be a candidate drug for antitumor immunotherapy.

Of the school-based factors that affect student performance, principal leadership is one of the most influential. Given the importance of school leadership to student success, leadership stability is also a [...]

Vesicular monoamine transporter 2 (VMAT2) accumulates monoamines in presynaptic vesicles for storage and exocytotic release, and has a vital role in monoaminergic neurotransmission1–3. Dysfunction of monoaminergic systems causes many neurological and psychiatric disorders, including Parkinson’s disease, hyperkinetic movement disorders and depression4–6. Suppressing VMAT2 with reserpine and tetrabenazine alleviates symptoms of hypertension and Huntington’s disease7,8, respectively. Here we describe cryo-electron microscopy structures of human VMAT2 complexed with serotonin and three clinical drugs at 3.5–2.8 Å, demonstrating the structural basis for transport and inhibition. Reserpine and ketanserin occupy the substrate-binding pocket and lock VMAT2 in cytoplasm-facing and lumen-facing states, respectively, whereas tetrabenazine binds in a VMAT2-specific pocket and traps VMAT2 in an occluded state. The structures in three distinct states also reveal the structural basis of the VMAT2 transport cycle. Our study establishes a structural foundation for the mechanistic understanding of substrate recognition, transport, drug inhibition and pharmacology of VMAT2 while shedding light on the rational design of potential therapeutic agents.

The corrosion mechanism of 510L low alloy steel treated by acid-cleaned surface (ACS) and eco-pickled surface (EPS) techniques in three simulated solutions (S0: atmospheric environment; S1: soil environment; S2: industrial environment) and the influence of interaction between different corrosive anions on corrosion were investigated. The results show that the total corrosion rates of samples in three simulated solutions were in order of S2 > S0 > S1, which is simultaneously correlated with initial corrosion dissolution processes as well as after the formation of corrosion products. HCO3−will inhibit the initial corrosion owing to the formation of films, whereas HSO3−will accelerate the dissolutions of the matrix based on the synergistic action of HSO3−and Cl−. On the other hand, there is no significant difference in corrosion rates between the samples treated by ACS and EPS techniques. The EPS technique that is safe, reusable and environmentally friendly can be further widely used in future work.

This study addresses the task assignment problem with robot coalitions, as encountered in practical scenarios, such as multiplayer reach-avoid games. Unlike the classical assignment problem where a single robot performs each task, the problem considered here involves tasks that require execution by a robot coalition consisting of two robots. This task assignment problem is a special instance of 3-set packing problem, which is known to be nondeterministic polynomial time (NP)-hard. We introduce the concept of $\epsilon$-coalition-competitive equilibrium ($\epsilon$-CCE) to characterize a kind of approximate solution that offers guaranteed performance. A distributed auction algorithm is developed to find an $\epsilon$-CCE within a finite number of iterations. In addition, several enhancements have been implemented to adapt the auction algorithm for practical applications where the task assignment problem may vary over time. Numerical simulations demonstrate that the distributed algorithm achieves satisfactory approximation quality.

•Testing the impact of different primer selections on fish diversity detection results.•Identifying the differences in primer selection monitoring results in different aquatic ecosystems. Emphasizing the superiority of the MiFish primer

This article studies a planar Homicidal Chauffeur reach-avoid differential game, where the pursuer is a Dubins car and the evader has simple motion. The pursuer aims to protect a goal region from the evader. The game is solved in an analytical approach instead of solving Hamilton–Jacobi–Isaacs equations numerically. First, an evasion region is introduced, based on which a pursuit strategy guaranteeing the winning of a simple-motion pursuer under specific conditions is proposed. Motivated by the simple-motion pursuer, a strategy for a Dubins-car pursuer is proposed when the pursuer–evader configuration satisfies separation condition (SC) and interception orientation (IO). The necessary and sufficient condition on capture radius, minimum turning radius, and speed ratio to guarantee the pursuit winning is derived. When the IO is deviated (Non-IO), a heading adjustment pursuit strategy is proposed, and the condition to achieve IO within a finite time is given. Based on it, a two-step pursuit strategy is proposed for the SC and Non-IO case. A nonconvex optimization problem is introduced to give a condition guaranteeing the winning of the pursuer. A polynomial equation gives a lower bound of the nonconvex problem, providing a sufficient and efficient pursuit winning condition. Finally, we extend to multiplayer games by collecting pairwise outcomes for pursuer–evader matchings. Simulations are provided to illustrate the theoretical results.

Thanks to their event-driven nature, spiking neural networks (SNNs) enjoy great energy efficiency and are believed to be a viable solution to the power wall problem. Due to difficulty in direct training, researchers have proposed indirect conversion-based methods, which allow SNNs to achieve comparable accuracy to their original nonspiking counterparts. However, most of these methods focus on classification problems; object detection with SNN, which involves a more challenging regression problem, remains an open research problem. The only existing SNN-based object detection method Spiking YOLO relied on a modified integrate-and-fire (IF) neuron model with two firing thresholds, which is not supported by most neuromorphic hardware. In order to run object detection on neuromorphic hardware, we propose a region-based SNN (R-SNN) with widely adopted IF neurons. The bounding box regressor uses mirror output neurons along with the original output neurons to represent both positive and negative bounding box offsets, and decodes the output spike trains with a simple bounding box recovery algorithm to recover real-valued bounding box offsets. Moreover, we deploy our R-SNN on our neuromorphic computing system “Darwin Mouse” to develop a low-power object detection application, demonstrating the feasibility of applying our method to real-world neuromorphic hardware with limited arithmetic precision. Experiments show that our R-SNN achieves a mean average precision (mAP) of 63.1% on VOC 2007, improving the state of the art (achieved by Spiking YOLO) by more than 11%.

The Transformer-based methods provide a good opportunity for modeling the global context of gigapixel whole slide image (WSI), however, there are still two main problems in applying Transformer to WSI-based survival analysis task. First, the training data for survival analysis is limited, which makes the model prone to overfitting. This problem is even worse for Transformer-based models which require large-scale data to train. Second, WSI is of extremely high resolution (up to 150,000 × 150,000 pixels) and is typically organized as a multi-resolution pyramid. Vanilla Transformer cannot model the hierarchical structure of WSI (such as patch cluster-level relationships), which makes it incapable of learning hierarchical WSI representation. To address these problems, in this article, we propose a novel Sparse and Hierarchical Transformer (SH-Transformer) for survival analysis. Specifically, we introduce sparse self-attention to alleviate the overfitting problem, and propose a hierarchical Transformer structure to learn the hierarchical WSI representation. Experimental results based on three WSI datasets show that the proposed framework outperforms the state-of-the-art methods.

Heteroatoms in situ-doped hierarchical porous hollow-activated carbons (HPHACs) have been prepared innovatively by pyrolyzation of setaria viridis combined with alkaline activation for the first time. The micro-morphology, pore structure, chemical compositions, and electrochemical properties are researched in detail. The obtained HPHACs are served as outstanding electrode materials in electrochemical energy storage ascribe to the particular hierarchical porous and hollow structure, and the precursor setaria viridis is advantage of eco-friendly as well as cost-effective. Electrochemical measurement results of the HPHACs electrodes exhibit not only high specific capacitance of 350 F g−1at 0.2 A g−1, and impressive surface specific capacitance (Cs) of 49.9 μF cm−2, but also substantial rate capability of 68% retention (238 F g−1at 10 A g−1) and good cycle stability with 99% retention over 5000 cycles at 5 A g−1in 6 M KOH. Besides, the symmetrical supercapacitor device based on the HPHACs electrodes exhibits excellent energy density of 49.5 Wh kg−1at power density of 175 W kg−1, but still maintains favorable energy density of 32.0 Wh kg−1at current density of 1 A g−1in 1-ethy-3-methylimidazolium tetrafluoroborate (EMIMBF4) ionic liquid electrolyte, and the excellent cycle stability behaviour shows the nearly 97% ratio capacitance retention of the initial capacitance after 10,000 cycles at current density of 2 A g−1. Overall, the results indicate that HPHACs derived from setaria viridis have appealing electrochemical performances thus are promising electrode materials for supercapacitor devices and large-scale applications.

Three-dimensional graphene–carbon nanotubes were prepared by the chemical reduction of a mixed aqueous solution of carboxylated multiwalled carbon nanotubes (C-MWCNTs) and graphene oxide, followed by freeze-drying. C-MWCNTs were then introduced into a three-dimensional graphene aerogel. The obtained three-dimensional graphene–carbon nanotubes were used as a solid-phase adsorbent for the determination of four organophosphorus pesticides (OPPs) in water using GC–MS. Under the optimal extraction conditions, good linearity was achieved between 0.5 and 500 μg L−1with a correlation coefficient of 0.9993–0.9998. The limits of detection (S/N= 3) and limits of quantification (S/N= 10) for the OPPs ranged from 0.28 to 0.52 and 0.96 to 1.64 μg L−1, respectively. The accuracy of the proposed method was evaluated by measuring the recovery of spiked samples, which ranged from 94.8 to 103.5%, with relative standard deviations lower than 7.0% (n= 3). This method was applied to the determination of target analytes in wetland water, lake water, and river water, and the results showed that the three-dimensional graphene–carbon nanotubes have great potential for the determination of OPPs in real water samples.

DNA cytosine methylation plays a vital role in repressing retrotransposons, and such derepression is linked with developmental failure, tumorigenesis and aging. DNA methylation patterns are formed by precisely regulated actions of DNA methylation writers (DNA methyltransferases) and erasers (TET, ten-eleven translocation dioxygenases). However, the mechanisms underlying target-specific oxidation of 5mC by TET dioxygenases remain largely unexplored. Here we show that a large low-complexity domain (LCD), located in the catalytic part of Tet enzymes, negatively regulates the dioxygenase activity. Recombinant Tet3 lacking LCD is shown to be hyperactive in converting 5mC into oxidized species in vitro. Endogenous expression of the hyperactive Tet3 mutant in mouse oocytes results in genome-wide 5mC oxidation. Notably, the occurrence of aberrant 5mC oxidation correlates with a consequent loss of the repressive histone mark H3K9me3 at ERVK retrotransposons. The erosion of both 5mC and H3K9me3 causes ERVK derepression along with upregulation of their neighboring genes, potentially leading to the impairment of oocyte development. These findings suggest that Tet dioxygenases use an intrinsic auto-regulatory mechanism to tightly regulate their enzymatic activity, thus achieving spatiotemporal specificity of methylome reprogramming, and highlight the importance of methylome integrity for development.

In this article, we study privacy-utility tradeoffs where users share privacy-correlated useful information with a service provider to obtain some utility. The service provider is adversarial in the sense that it can infer the users' private information based on the shared useful information. To minimize the privacy leakage while maintaining a desired level of utility, the users carefully perturb the useful information via a probabilistic privacy mapping before sharing it. We focus on the setting in which the adversary attempting an inference attack on the users' privacy has biased information about the statistical correlation between the private and useful variables. This information asymmetry between the users and the limited adversary is shown to lead to better privacy guarantees. We first identify assumptions on the adversary's information so that the inference costs are well-defined. Then, we characterize the impact of the information asymmetry and show that it increases the inference costs for the adversary. We further formulate the design of the privacy mapping against a limited adversary as a difference of convex functions program and solve it via the concave–convex procedure. When the adversary's information is not precisely available, we adopt a Bayesian view and represent the adversary's information by a probability distribution. In this case, the expected cost for the adversary does not admit a closed-form expression, and we establish and maximize a lower bound of it. We provide a numerical example to illustrate the theoretical results.

This article considers a planar multiple-pursuer single-evader reach-avoid game played in a constant flow field. Such a game problem arises in practical applications, such as maritime escort, aircraft confrontation in airflow fields, and defense of moving targets. The dominance boundary, which serves as an analogy to the Apollonius circle and the Cartesian oval, is investigated. The existence of semipermeable surfaces in the game is shown. Under reasonable assumptions, there exists a semipermeable surface, which is part of the barrier surface separating the winning regions for the evader and pursuers. The winning strategies are designed for both sides of players in their corresponding winning regions. It is also shown that at most two pursuers are necessary for successful protection, as in the situation where there is no flow field. Finally, numerical results are given to illustrate the conclusions.

This article examined the impact of the supply chain specific investments (SCSIs) of small and medium-sized enterprises (SMEs) on their enterprise performance (EP). In addition, the mediating role of supply chain financing capability (SCFC) and the moderating role of digital technology adoption (DTA) were also investigated. The valid representative sample for this study was drawn from 380 Chinese SMEs. Multiple regression analysis and a structural equation model were used to test the proposed theoretical model and research hypotheses. The analysis illustrated the positive influence of SCSIs (supply chain asset-specific investments and supply chain relationship-specific investments) of SMEs on their EP. Furthermore, SCFC was found to mediate the relationship between SCSIs and EP, while DTA positively moderated the relationship between SCSIs and SCFC of SMEs. To the best of our knowledge, this is the first study to explore the impact of SCSIs on EP of SMEs based on resource orchestration theory. In addition, this study innovatively sheds light on the internal mechanism and contextual mechanism on the effectiveness of SCSIs for SMEs, through the introduction of two factors, namely SCFC and DTA.

Aim: To investigate the social contact patterns of Chinese residents in the post COVID-19 pandemic period, explore the associations between the different contact patterns and incidence of influenza. Subject and methods: A cross-sectional social contact survey was conducted on Chinese residents from December 9, 2021, to January 20, 2022. The contact was divided into participant–acquaintance (P–A) and participant–stranger (P–S), and the generalized linear model (GLM) was used to explore their impact on the incidence of influenza. Results: The contact frequency of P–S was higher than that of P–A (3.41 V.S. 1.94), but the average contact duration of P–S was much shorter than that of P–A (0.20 h V.S. 5.53 h). P–A contact mainly occurred among peers as well as family members. However, P–S contact was concentrated in the middle-aged population. In GLM, the influence of P–S on the incidence of influenza was higher (OR = 2.63, 95% CI 2.29–3.02) than that of P–A (OR = 1.82, 95% CI 1.56–2.13), especially on the weekend (P–S OR = 4.31, 95% CI 2.31–8.01) and in recreation places (P–S OR = 4.94, 95% CI 2.26–10.78). Conclusions: Social contact will increase the incidence rate of influenza. Therefore, when contacting with strangers, especially in recreation places, personal protective behavior should be strengthened.

Photocatalytic degradation is a promising method for controlling the increasing contamination of the water environment due to pharmacologically active compounds (PHACs). Herein, oxygen vacancy (OV)-modulated Z-scheme CuWO4/CuBi2O4hybrid systems were fabricated viathermal treatment by loading of CuWO4nanoparticles with OVs on CuBi2O4surfaces. The synthesized CuWO4/CuBi2O4hybrid samples exhibited an enhanced photodegradation ability to remove PHACs under visible-light irradiation. More importantly, an optimized sample (10 wt % CuWO4/CuBi2O4) exhibited superior catalytic activity and excellent recycling stability for PHAC photodegradation. In addition, possible degradation paths for PHAC removal over the CuWO4/CuBi2O4hybrid systems were proposed. The enhanced photocatalytic performance could be attributed to the efficient separation and transfer of photoformed charge pairs viathe Z-scheme mechanism. This Z-scheme mechanism was systematically analyzed using trapping experiments of active species, ultraviolet photoelectron spectroscopy, electron spin resonance, and the photodepositions of noble metals. The findings of this study can pave the way for developing highly efficient Z-scheme photocatalytic systems for PHAC photodegradation.

The natural locomotion interface is critical to the development of many VR applications. For household VR applications, there are two basic requirements: natural immersive experience and minimized space occupation. The existing locomotion strategies generally do not simultaneously satisfy these two requirements well. This article presents a novel omnidirectional treadmill (ODT) system named Hex-Core-MK1 (HCMK1). By implementing two kinds of mirror-symmetrical spiral rollers to generate the omnidirectional velocity field, this proposed system is capable of providing real walking experiences with a full-degree of freedom in an area as small as 1.76 m$^{2}$2, while delivering great advantages over several existing ODT systems in terms of weight, volume, latency and dynamic performance. Compared with the sizes of Infinadeck and HCP, the two best motor-driven ODTs so far, the 8 cm height of HCMK1 is only 20% of Infinadeck and 50% of HCP. In addition, HCMK1 is a lightweight device weighing only 110 kg, which provides possibilities for further expanding VR scenarios, such as terrain simulation. The system latency of HCMK1 is only 9ms. The experiments show that HCMK1 can deliver a starting acceleration of 16.00 m/s$^{2}$2 and a braking acceleration of 30.00 m/s$^{2}$2.

Spatio-temporal action localization aims to locate the spatial and temporal positions of actors and classify their actions. However, prior research overlooks the fact that human actions often interact with novel objects in real-world scenarios, which neglects the various combinations of action-object, and considerably limits the generalization of the developed models. In this paper, we study the action-object combinations by researching multi-modal vision information of them. To this end, we propose a novel compositional spatio-temporal action localization (Com-STAL) task, which features non-overlapping action-object combinations in their training and test sets. Based on this, we construct a compositional action localization dataset (Com-AD). Beyond that, we propose a simple yet effective framework, Instance-Centric Interaction Network (ICIN), to reduce invalid induction biases within the visual modality and alleviate the combined distribution bias issue by leveraging additional modal information. The extensive experiment results on Com-AD demonstrate superior action localization performance of ICIN.

Survival analysis is a significant study in cancer prognosis, and the multi-modal data, including histopathological images, genomic data, and clinical information, provides unprecedented opportunities for its development. However, because of the high dimensionality and the heterogeneity of histopathological images and genomic data, acquiring effective predictive characters from these multi-modal data has always been a challenge for survival analysis. In this article, we propose a transformer-based survival analysis model (TransSurv) for colorectal cancer that can effectively integrate intra-modality and inter-modality features of histopathological images, genomic data, and clinical information. Specifically, to integrate the intra-modality relationship of image patches, we develop a multi-scale histopathological features fusion transformer (MS-Trans). Furthermore, we provide a cross-modal fusion transformer based on cross attention for multi-scale pathological representation and multi-omics representation, which includes RNA-seq expression and copy number alteration (CNA). At the output layer of the TransSurv, we adopt the Cox layer to integrate multi-modal fusion representation with clinical information for end-to-end survival analysis. The experimental results on the Cancer Genome Atlas (TCGA) colorectal cancer cohort demonstrate that the proposed TransSurv outperforms the existing methods and improves the prognosis prediction of colorectal cancer.

Publishing the research works on academic publications is an important part of the scientific process. Since the development of computer science research is very fast, researchers tend to publish the research works in a fast way, such as conferences whose review processes are faster than the journals. In the past decades, one conference usually focuses on a specific research field and the topic or method overlap between conferences is low. We have noticed that, in recent years, some topics or methods which were once studied in a small number of specific research fields have become popular in many other fields. Naturally, we come up with two research questions: (1) Do the conferences indeed become similar? and (2) How do conferences become similar? In this paper, we first use a trans-temporal heterogeneous graph network to model academic conferences in recent 20 years. Due to the large number of conferences, we categorize these conferences into 6 research fields for brevity. Then, we first quantitatively and qualitatively assess “Do the research fields become similar?” and then focus on exploring “How do research fields become similar?”. From the result, we find the reason for the research fields in computer science become similar is that AI becomes pervasive and researchers tend to apply the machine learning methods to different application fields. Since the methods become universal between different research fields, researchers should pay more attention to advanced information in other fields to motivate more interdisciplinary works. To assist the researchers to explore related interdisciplinary advanced information, it is crucial to measure the cross-field impact of papers using the citation information and recommend the paper which has a high cross-field impact on the related researchers. As for the newly published papers which do not have any citations, we also propose a cross-field impact prediction model to recommend the cutting-edge research works to related researchers accurately. Experiments conducted on real-world datasets verify the effectiveness of the proposed method.

Multiturn dialogue systems with cognitive capabilities have been popularly used in many applications and generate a large volume of dialogue texts. The dialogue texts of many dialogue systems contain multiple labels, and a typical case is that the customer requirements or complaints in an online multiturn dialogue service system often need to be classified into multiple categories or forwarded to multiple departments, that is, multilabel classification on multiturn dialogue texts. Traditional methods treat the dialogue text as the plain text, which makes it difficult for classifiers to learn important features. On the other hand, regarding each utterance as the basic semantic unit may destroy the semantic information integrity of the dialogue. Furthermore, the above methods all ignore the relationship between labels. To address above issues, we propose to use the utterance pair as the basic semantic unit and design a novel multilabel classification model for multiturn dialogue texts in this article. Specifically, we propose to use the multihead attention mechanism to learn the key features of multiple labels and design a new method to learn the correlation among labels. Besides, we propose to use the penalty term to allow multihead attention to focus on different semantics. Finally, we conduct sufficient experiments on four real-world multiturn dialogue data sets and the experimental results show that our model produces superior performance.

Reliability is an important metric for unmanned aerial vehicles (UAVs) to perform important, complex, and dangerous tasks. In addition, reliability influences the operational cost in UAVs routing. In order to reduce the operational cost of UAVs tasks, a method of routing strategy optimization is proposed from a view of mission reliability in this paper to address the limitations in routing scheduling. Different from previous studies on the UAVs routing optimization problem, this paper proposes a method that can reduce the operational costs of UAVs tasks based on the mission reliability. This method includes two stages: the pre-optimization stage and the rescheduling stage. In the pre-optimization stage, an optimal UAVs route solution is obtained for all the targets, while in the rescheduling stage, new UAVs are dispatched to the unvisited targets in the pre-optimization stage based on the new optimal UAVs route.

Cancer immune escape, metastasis, recurrence, and multidrug resistance are all associated with hypoxia in the tumor microenvironment (TME). We synthesized a CuPPaCC conjugate for reactive oxygen species (ROS)-mediated cancer therapy. CuPPaCC continuously produced cytotoxic ROS and oxygen through a photo-chemocycloreaction, alleviated hypoxia, and inhibited the expression of a hypoxia-inducing factor (HIF-1α). CuPPaCC was synthesized from pyromania phyllophyllic acid a (PPa), cystine (CC), and copper ions, and its structure was characterized by nuclear magnetic resonance (NMR) and mass spectrometry (MS). The ability of CuPPaCC to produce ROS and oxygen after photodynamic therapy (PDT) in vitro and in vivo was investigated. The ability of CuPPaCC to consume glutathione was investigated. CuPPaCC toxicity (light and dark) in CT26 cells was analyzed by MTT and live/dead cell staining. The anticancer effect of CuPPaCC in vivo was investigated in CT26 Balb/c mice. When stimulated by the TME, CuPPaCC released Cu2+and PPaCC, and the singlet oxygen yield increased from 34 to 56.5%. The dual ROS-generating mechanism via a Fenton-like reaction/photoreaction and dual glutathione depletion via Cu2+/CC multiplied the antitumor efficacy of CuPPaCC. The photo-chemocycloreaction continued to produce oxygen and maintained high ROS levels even after PDT, significantly alleviating hypoxia in the TME and downregulating the expression of HIF-1α. CuPPaCC thus showed excellent antitumor activity in vitro and in vivo. These results showed that the strategy could be effective in improving the antitumor efficacy of CuPPaCC and could be used as a synergistic regimen for cancer therapy.

By using the central composite design based on the response surface method, the multiobjective collaborative optimization of Fe, Cr and Ni metal recovery from stainless steel dust and laterite nickel ore was achieved through carbothermal reduction, and a high metal recovery rate and high-grade Fe-Cr-Ni-C alloy particles were obtained. The addition of laterite nickel ore and reduction temperature exert significant effects on the metal recovery, and a significant interaction is observed among the recoveries of the metals Fe, Cr and Ni in the reduction products. The optimal process parameters obtained through the optimization of the model are listed as follows: the amount of added laterite nickel ore is 5.47%, the reduction temperature is 1428.02 °C, the reduction time is 23.10 min, and the carbon oxygen ratio (FC/O) is 0.85. The predicted results for Fe, Cr and Ni recovery using the model are 93.15%, 91.63% and 92.70%, respectively.

Short peptides that can respond to external stimuli have been considered as the preferred building blocks to construct hydrogels for biomedical applications. In particular, photoresponsive peptides that are capable of triggering the formation of hydrogels upon light irradiation allow the properties of hydrogels to be changed remotely by precise and localized actuation. Here, we used the photochemical reaction of the 2-nitrobenzyl ester group (NB) to develop a facile and versatile strategy for constructing photoactivated peptide hydrogels. The peptides with high aggregation propensity were designed as hydrogelators, which were photocaged by a positively charged dipeptide (KK) to provide strong charge repulsion and prevent self-assembly in water. Light irradiation led to the removal of KK and triggered the self-assembly of peptides and the formation of hydrogel. Light stimulation endows spatial and temporal control, which enables the formation of hydrogel with precisely tunable structure and mechanical properties. Cell culture and behavior study indicated that the optimized photoactivated hydrogel was suitable for 2D and 3D cell culture, and its photocontrollable mechanical strength could regulate the spreading of stem cells on its surface. Therefore, our strategy provides an alternative way to construct photoactivated peptide hydrogels with wide applications in biomedical areas.

Using single-molecule displacement/diffusivity mapping (SMdM), an emerging super-resolution microscopy method, here we quantify, at nanoscale resolution, the diffusion of a typical fluorescent protein (FP) in the endoplasmic reticulum (ER) and mitochondrion of living mammalian cells. We thus show that the diffusion coefficients Din both organelles are ∼40% of that in the cytoplasm, with the latter exhibiting higher spatial inhomogeneities. Moreover, we unveil that diffusions in the ER lumen and the mitochondrial matrix are markedly impeded when the FP is given positive, but not negative, net charges. Calculation shows most intraorganellar proteins as negatively charged, hence a mechanism to impede the diffusion of positively charged proteins. However, we further identify the ER protein PPIB as an exception with a positive net charge and experimentally show that the removal of this positive charge elevates its intra-ER diffusivity. We thus unveil a sign-asymmetric protein charge effect on the nanoscale intraorganellar diffusion.

The flagella of mammalian sperm display non-planar, asymmetric beating, in contrast to the planar, symmetric beating of flagella from sea urchin sperm and unicellular organisms. The molecular basis of this difference is unclear. Here, we perform in situ cryo-electron tomography of mouse and human sperm, providing the highest-resolution structural information to date. Our subtomogram averages reveal mammalian sperm-specific protein complexes within the microtubules, the radial spokes and nexin–dynein regulatory complexes. The locations and structures of these complexes suggest potential roles in enhancing the mechanical strength of mammalian sperm axonemes and regulating dynein-based axonemal bending. Intriguingly, we find that each of the nine outer microtubule doublets is decorated with a distinct combination of sperm-specific complexes. We propose that this asymmetric distribution of proteins differentially regulates the sliding of each microtubule doublet and may underlie the asymmetric beating of mammalian sperm.

On May 21, a Ms6.4 earthquake occurred in Yangbi, Yunnan Province. Less than 5 ​h later, on May 21, follows a Ms7.4 earthquake in Madoi, Qinghai Province. In this study, multiple parameters are studied from CSES-01 satellite and GNSS. The parameters include the electron density from LAP, energetic particle flux from HEPP, electric field spectrum from EFD, induced magnetic field spectrum from SCM, and global TEC from ground-based GNSS receivers. In the time scale, all parameters changed on May 15, 17 and 18. The relative amplitude of Ne is more than 200%. In space scale, the clear enhancement focus on the southeast of the epicenter of Madoi, Qinghai, and the shape of the focus region is like a clockwise-30°-rotated rectangle, which is about 45° in length and 20° in width. Moreover, there appeared enhancement in the conjugated region. According to the day-to-day analysis of the low-frequency (75–90 Hz) spectrum of EFD and SCM payloads, the relationship of the increment on May 18 could be confirmed with the successive two earthquakes. The energetic particle fluxes enhanced on May 20. All those changes are located in the east of the following epicenter and appeared just several days before, which could be well explained by the Lithosphere-Atmosphere-Ionosphere theory.

Soil microbial community composition is sensitive to environmental changes and can be used as a predictor of grassland degradation. However, the response of microbial communities to alpine grassland degradation has not been well elucidated. Illumina MiSeq high-throughput sequencing technology was used to analyze the composition and diversity of soil microbial communities in three degraded grassland types, namely, non-degradation, moderate degradation, and severe degradation, in the alpine grasslands of Qilian Mountains. With the deterioration of alpine grasslands, soil pH increased significantly (P< 0.01). Soil water content (SWC), soil organic matter (SOC), total nitrogen (TN), total phosphorus (TP), available nitrogen (AN), and available phosphorus (AP) showed a significant decreasing trend (P< 0.01). Soil microbial biomass carbon (MBC), microbial biomass nitrogen (MBN), and microbial biomass phosphorus (MBP) significantly decreased (P< 0.01). Alpine grassland degradation significantly changed the community composition, and β-diversity of soil bacteria and fungi changed significantly, but their α-diversity did not change significantly. Soil pH, SWC, AN, and MBN were the key factors limiting the changes in soil bacterial and fungal community composition. Alpine grassland degradation has changed soil microbial community composition, which should be the future focus of alpine grassland degradation restoration management.

The chromatin-modifying enzyme ATAD2 confers oncogenic competence and proliferative advantage in malignances. We previously identified ATAD2 as a marker and driver of cell proliferation in ovarian cancer (OC); however, the mechanisms whereby ATAD2 is regulated and involved in cell proliferation are still unclear. Here, we disclose that ATAD2 displays a classical G2/M gene signature, functioning to facilitate mitotic progression. ATAD2 ablation caused mitotic arrest and decreased the ability of OC cells to pass through nocodazole-arrested mitosis. ChIP-seq data analyses demonstrated that DREAM and MYBL2-MuvB (MMB), two switchable MuvB-based complexes, bind the CHR elements in the ATAD2 promoter, representing a typical feature and principle mechanism of the periodic regulation of G2/M genes. As a downstream target of MYBL2, ATAD2 deletion significantly impaired MYBL2-driven cell proliferation. Intriguingly, ATAD2 silencing also fed back to destabilize the MYBL2 protein. The significant coexpression of MYBL2 and ATAD2 at both the bulk tissue and single-cell levels highlights the existence of the MYBL2-ATAD2 signaling in OC patients. This signaling is activated during tumorigenesis and correlated with TP53mutation, and its hyperactivation was found especially in high-grade serous and drug-resistant OCs. Disrupting this signaling by CRISPR/Cas9-mediated ATAD2 ablation inhibited the in vivo growth of OC in a subcutaneous tumor xenograft mouse model, while pharmacologically targeting this signaling with an ATAD2 inhibitor demonstrated high therapeutic efficacy in both drug-sensitive and drug-resistant OC cells. Collectively, we identified a novel MYBL2-ATAD2 proliferative signaling axis and highlighted its potential application in developing new therapeutic strategies, especially for high-grade serous and drug-resistant OCs.

Group activity recognition (GAR) is a challenging task aimed at recognizing the behavior of a group of people. It is a complex inference process in which visual cues collected from individuals are integrated into the final prediction, being aware of the interaction between them. This paper goes one step further beyond the existing approaches by designing a Hierarchical Graph-based Cross Inference Network (HiGCIN), in which three levels of information, i.e., the body-region level, person level, and group-activity level, are constructed, learned, and inferred in an end-to-end manner. Primarily, we present a generic Cross Inference Block (CIB), which is able to concurrently capture the latent spatiotemporal dependencies among body regions and persons. Based on the CIB, two modules are designed to extract and refine features for group activities at each level. Experiments on two popular benchmarks verify the effectiveness of our approach, particularly in the ability to infer with multilevel visual cues. In addition, training our approach does not require individual action labels to be provided, which greatly reduces the amount of labor required in data annotation.

The collection and curation of large-scale medical datasets from multiple institutions is essential for training accurate deep learning models, but privacy concerns often hinder data sharing. Federated learning (FL) is a promising solution that enables privacy-preserving collaborative learning among different institutions, but it generally suffers from performance deterioration due to heterogeneous data distributions and a lack of quality labeled data. In this paper, we present a robust and label-efficient self-supervised FL framework for medical image analysis. Our method introduces a novel Transformer-based self-supervised pre-training paradigm that pre-trains models directly on decentralized target task datasets using masked image modeling, to facilitate more robust representation learning on heterogeneous data and effective knowledge transfer to downstream models. Extensive empirical results on simulated and real-world medical imaging non-IID federated datasets show that masked image modeling with Transformers significantly improves the robustness of models against various degrees of data heterogeneity. Notably, under severe data heterogeneity, our method, without relying on any additional pre-training data, achieves an improvement of 5.06%, 1.53% and 4.58% in test accuracy on retinal, dermatology and chest X-ray classification compared to the supervised baseline with ImageNet pre-training. In addition, we show that our federated self-supervised pre-training methods yield models that generalize better to out-of-distribution data and perform more effectively when fine-tuning with limited labeled data, compared to existing FL algorithms. The code is available at https://github.com/rui-yan/SSL-FL.

Carbon dots (CDs) have great potential for cancer diagnosis and treatment. Photodynamic therapy and chemodynamic therapy are promising treatments mediated by reactive oxygen species (ROS), which have the advantages of being minimally invasive, having no multi-drug resistance, and having no systemic toxic side effects. However, the tumor microenvironment (TME) and poor targetability often reduce the therapeutic effect. In this work, we have successfully prepared folate-based carbon dots (FCP-CDs) from folic acid (FA), citric acid (CA), and polyethyleneimine (PEI) for tumor-targeting. The surface of FCP-CDs was modified using organic disulfide, 3,3′-dithiodipropionic acid (DTPA), and a photosensitizer (PS) pyropheophorbide-a (PPa) to form a tumor microenvironment-responsive nanoplatform, FCP-CDs@DTPA@PPa (named FCPPD), for synergistic cancer therapy. The results showed that FCPPD effectively preserved the tumor target specificity of folic acid and the photodynamic therapeutic (PDT) activity of PPa, and could provide additional chemodynamic therapeutic (CDT) function by reacting with hydrogen peroxide (H2O2) to generate OH. The introduction of DTPA, which contains disulfide bonds, endows FCPPD with an excellent ability to deplete glutathione (GSH) in tumors viaintracellular redox reactions, amplifying intracellular oxidative strain and enhancing ROS-based therapeutic effects. Systematic in vitroand in vivostudies under various conditions have shown that the obtained FCPPD nanoparticles have good biocompatibility and could be a promising therapeutic agent for imaging-guided PDT/CDT combination therapy.

Backpropagation has been successfully generalized to optimize deep spiking neural networks (SNNs), where, nevertheless, gradients need to be propagated back through all layers, resulting in a massive consumption of computing resources and an obstacle to the parallelization of training. A biologically motivated scheme of local learning provides an alternative to efficiently train deep networks but often suffers a low performance of accuracy on practical tasks. Thus, how to train deep SNNs with the local learning scheme to achieve both efficient and accurate performance still remains an important challenge. In this study, we focus on a supervised local learning scheme where each layer is independently optimized with an auxiliary classifier. Accordingly, we first propose a spike-based efficient local learning rule by only considering the direct dependencies in the current time. We then propose two variants that additionally incorporate temporal dependencies through a backward and forward process, respectively. The effectiveness and performance of our proposed methods are extensively evaluated with six mainstream datasets. Experimental results show that our methods can successfully scale up to large networks and substantially outperform the spike-based local learning baselines on all studied benchmarks. Our results also reveal that gradients with temporal dependencies are essential for high performance on temporal tasks, while they have negligible effects on rate-based tasks. Our work is significant as it brings the performance of spike-based local learning to a new level with the computational benefits being retained.