We overview several properties—old and new—of training overparameterized deep networks under the square loss. We first consider a model of the dynamics of gradient flow under the square loss in deep homogeneous rectified linear unit networks. We study the convergence to a solution with the absolute minimum ρ, which is the product of the Frobenius norms of each layer weight matrix, when normalization by Lagrange multipliers is used together with weight decay under different forms of gradient descent. A main property of the minimizers that bound their expected error for a specific network architecture is ρ. In particular, we derive novel norm-based bounds for convolutional layers that are orders of magnitude better than classical bounds for dense networks. Next, we prove that quasi-interpolating solutions obtained by stochastic gradient descent in the presence of weight decay have a bias toward low-rank weight matrices, which should improve generalization. The same analysis predicts the existence of an inherent stochastic gradient descent noise for deep networks. In both cases, we verify our predictions experimentally. We then predict neural collapse and its properties without any specific assumption—unlike other published proofs. Our analysis supports the idea that the advantage of deep networks relative to other classifiers is greater for problems that are appropriate for sparse deep architectures such as convolutional neural networks. The reason is that compositionally sparse target functions can be approximated well by “sparse” deep networks without incurring in the curse of dimensionality.

Cancer treatments always face challenging problems, particularly drug resistance due to tumor cell heterogeneity. The existing datasets include the relationship between gene expression and drug sensitivities; however, the majority are based on tissue-level studies. Study drugs at the single-cell level are perspective to overcome minimal residual disease caused by subclonal resistant cancer cells retained after initial curative therapy. Fortunately, machine learning techniques can help us understand how different types of cells respond to different cancer drugs from the perspective of single-cell gene expression. Good modeling using single-cell data and drug response information will not only improve machine learning for cell–drug outcome prediction but also facilitate the discovery of drugs for specific cancer subgroups and specific cancer treatments. In this paper, we review machine learning and deep learning approaches in drug research. By analyzing the application of these methods on cancer cell lines and single-cell data and comparing the technical gap between single-cell sequencing data analysis and single-cell drug sensitivity analysis, we hope to explore the trends and potential of drug research at the single-cell data level and provide more inspiration for drug research at the single-cell level. We anticipate that this review will stimulate the innovative use of machine learning methods to address new challenges in precision medicine more broadly.

Host defense peptides (HDPs) are one of the potentially promising agents for infection diseases due to their broad spectrum and low resistance rate, but their clinical applications are limited by proteolytic instability, high-cost, and complicated synthesis process. Here, we report a host-defense-peptide-mimicking β-peptide polymer that resists proteolysis to have enhanced the activity under physiological conditions, excellent antimicrobial efficiency even at high density of bacteria, and low cost for preparation. The β-peptide polymer demonstrated quorum sensing (QS) interference and bactericidal effect against both bacterial communities and individual bacterium to simultaneously block bacterial communication and disrupt bacterial membranes. The hierarchical QS network was suppressed, and main QS signaling systems showed considerably down-regulated gene expression, resulting in excellent biofilm eradication and virulence reduction effects. The dual-modal antibacterial ability possessed excellent therapeutic effects in Pseudomonas aeruginosa pneumonia, which could inhibit biofilm formation and exhibit better antibacterial and anti-inflammatory efficiency than clinically used antibiotics, levofloxacin. Furthermore, the β-peptide polymer also showed excellent therapeutic effect Escherichia coli pyogenic liver abscess. Together, we believed that the β-peptide polymer had a feasible clinical potential to treat bacterial infection diseases.

As a naturally occurring cytolytic peptide, melittin (Mel) has strong cytolytic activity and is a potent therapeutic peptide for cancer therapy. However, the serious hemolytic activity of Mel largely impedes its clinical applications. In this work, based on the strong interactions between proteins/peptides and polyphenols, we develop a tannic acid–Fe³⁺ metal–phenolic network (MPN)-based strategy that can convert Mel from foe to friend via shielding its positive charges and reducing its hemolytic activity. Besides, an immune adjuvant resiquimod (R848) is also introduced for immunostimulation, affording the final Mel- and R848-coloaded nanodrug. The Mel-caused membrane disruption can induce immunogenic cell death for immunostimulation, R848 can act as an immune adjuvant to further facilitate the immunostimulatory effect, and the tannic acid–Fe³⁺ MPN-mediated Fenton reaction can produce reactive oxygen species for cancer treatment. Further experiments reveal that the nanodrug can effectively cause immunogenic cell death of tumor cells and arouse robust intratumoral and systemic antitumor immunostimulation. In the bilateral tumor-bearing mouse models, the nanodrug considerably destroys the primary tumor and also boosts the abscopal effect to ablate the distant tumor. Collectively, the MPN-facilitated “foe-to-friend” strategy may promote the practical applications of Mel and foster the development of cancer immunotherapeutics.

The design of new materials with special performances is still a great challenge, especially for the deep-ultraviolet nonlinear optical materials in which it is difficult to balance large bandgaps and strong second harmonic generation responses due to their inverse relationship. Cation variation not only influences the whole structure frameworks but also directly participates in the formation of electronic structures, both of which could lead to the uncontrollability of the properties of the designed materials. Here, a novel approach, aiming at purposeful and foreseeable material designs, is proposed to characterize the role of cations. By the verification of several series of borates, the influences of cation variation on property changes are explored systematically. Accordingly, a feasible strategy of designing deep-ultraviolet nonlinear optical materials by substituting barium for lead has been concluded, which could obviously blue-shift the ultraviolet cutoff edge and maintain the relatively strong second harmonic generation response (more than 2 times of KH₂PO₄), achieving the property optimization, and especially works efficiently in fluorooxoborates. The property optimization design strategy and the cation characterization method are not only helpful in exploring nonlinear optical materials but also enlightening in material design and selection.

Efficient in vivo delivery of anti-inflammatory proteins to modulate the microenvironment of an injured spinal cord and promote neuroprotection and functional recovery is a great challenge. Nucleoside-modified messenger RNA (mRNA) has become a promising new modality that can be utilized for the safe and efficient delivery of therapeutic proteins. Here, we used lipid nanoparticle (LNP)-encapsulated human interleukin-10 (hIL-10)-encoding nucleoside-modified mRNA to induce neuroprotection and functional recovery following rat spinal cord contusion injury. Intralesional administration of hIL-10 mRNA-LNP to rats led to a remarkable reduction of the microglia/macrophage reaction in the injured spinal segment and induced significant functional recovery compared to controls. Furthermore, hIL-10 mRNA treatment induced increased expression in tissue inhibitor of matrix metalloproteinase 1 and ciliary neurotrophic factor levels in the affected spinal segment indicating a time-delayed secondary effect of IL-10 5 d after injection. Our results suggest that treatment with nucleoside-modified mRNAs encoding neuroprotective factors is an effective strategy for spinal cord injury repair.

Over the past 60 years, the semiconductor industry has been the core driver for the development of information technology, contributing to the birth of integrated circuits, Internet, artificial intelligence, and Internet of Things. Semiconductor technology has been evolving in structure and material with co-optimization of performance–power–area–cost until the state-of-the-art sub-5-nm node. Two-dimensional (2D) semiconductors are recognized by the industry and academia as a hopeful solution to break through the quantum confinement for the future technology nodes. In the recent 10 years, the key issues on 2D semiconductors regarding material, processing, and integration have been overcome in sequence, making 2D semiconductors already on the verge of application. In this paper, the evolution of transistors is reviewed by outlining the potential of 2D semiconductors as a technological option beyond the scaled metal oxide semiconductor field-effect transistors. We mainly focus on the optimization strategies of mobility (μ), equivalent oxide thickness (EOT), and contact resistance (R_C), which enables high ON current (I_on) with reduced driving voltage (V_dd). Finally, we prospect the semiconductor technology roadmap by summarizing the technological development of 2D semiconductors over the past decade.

The lack of efficient crop phenotypic measurement methods has become a bottleneck in the field of breeding and precision cultivation. However, high-throughput and accurate phenotypic measurement could accelerate the breeding and improve the existing cultivation management technology. In view of this, this paper introduces a high-throughput crop phenotype measurement platform named the LQ-FieldPheno, which was developed by China National Agricultural Information Engineering Technology Research Centre. The proposed platform represents a mobile phenotypic high-throughput automatic acquisition system based on a field track platform, which introduces the Internet of Things (IoT) into agricultural breeding. The proposed platform uses the crop phenotype multisensor central imaging unit as a core and integrates different types of equipment, including an automatic control system, upward field track, intelligent navigation vehicle, and environmental sensors. Furthermore, it combines an RGB camera, a 6-band multispectral camera, a thermal infrared camera, a 3-dimensional laser radar, and a deep camera. Special software is developed to control motions and sensors and to design run lines. Using wireless sensor networks and mobile communication wireless networks of IoT, the proposed system can obtain phenotypic information about plants in their growth period with a high-throughput, automatic, and high time sequence. Moreover, the LQ-FieldPheno has the characteristics of multiple data acquisition, vital timeliness, remarkable expansibility, high-cost performance, and flexible customization. The LQ-FieldPheno has been operated in the 2020 maize growing season, and the collected point cloud data are used to estimate the maize plant height. Compared with the traditional crop phenotypic measurement technology, the LQ-FieldPheno has the advantage of continuously and synchronously obtaining multisource phenotypic data at different growth stages and extracting different plant parameters. The proposed platform could contribute to the research of crop phenotype, remote sensing, agronomy, and related disciplines.

Locomotion in mollusc Aplysia is implemented by a pedal rolling wave, a type of axial locomotion. Well-studied examples of axial locomotion (pedal waves in Drosophila larvae and body waves in leech, lamprey, and fish) are generated in a segmented nervous system via activation of multiple coupled central pattern generators (CPGs). Pedal waves in molluscs, however, are generated by a single pedal ganglion, and it is unknown whether there are single or multiple CPGs that generate rhythmic activity and phase shifts between different body parts. During locomotion in intact Aplysia, bursting activity in the parapedal commissural nerve (PPCN) was found to occur during tail contraction. A cluster of 20 to 30 P1 root neurons (P1Ns) on the ventral surface of the pedal ganglion, active during the pedal wave, were identified. Computational cluster analysis revealed that there are 2 phases to the motor program: phase I (centered around 168°) and phase II (centered around 357°). PPCN activity occurs during phase II. The majority of P1Ns are motoneurons. Coactive P1Ns tend to be electrically coupled. Two classes of pedal interneurons (PIs) were characterized. Class 1 (PI1 and PI2) is active during phase I. Their axons make a loop within the pedal ganglion and contribute to locomotor pattern generation. They are electrically coupled to P1Ns that fire during phase I. Class 2 (PI3) is active during phase II and innervates the contralateral pedal ganglion. PI3 may contribute to bilateral coordination. Overall, our findings support the idea that Aplysia pedal waves are generated by a single CPG.

Microglia are resident macrophage cells in the central nervous system that search for pathogens or abnormal neural activities and migrate to resolve the issues. The effective search and targeted motion of macrophages mean dearly to maintaining a healthy brain, yet little is known about their migration dynamics. In this work, we study microglial motion with and without the presence of external mechanostimuli. We discover that the cells are promptly attracted by the applied forces (i.e., mechanotaxis), which is a tactic behavior as yet unconfirmed in microglia. Meanwhile, in both the explorative and the targeted migration, microglia display dynamics that is strikingly analogous to bacterial run-and-tumble motion. A closer examination reveals that microglial run-and-tumble is more sophisticated, e.g., they display a short-term memory when tumbling and rely on active steering during runs to achieve mechanotaxis, probably via the responses of mechanosensitive ion channels. These differences reflect the sharp contrast between microglia and bacteria cells (eukaryotes vs. prokaryotes) and their environments (compact tissue vs. fluid). Further analyses suggest that the reported migration dynamics has an optimal search efficiency and is shared among a subset of immune cells (human monocyte and macrophage). This work reveals a fruitful analogy between the locomotion of 2 remote systems and provides a framework for studying immune cells exploring complex environments.

In recent years, brain science and neuroscience have greatly propelled the innovation of computer science. In particular, knowledge from the neurobiology and neuropsychology of the brain revolutionized the development of reinforcement learning (RL) by providing novel interpretable mechanisms of how the brain achieves intelligent and efficient decision making. Triggered by this, there has been a boom in research about advanced RL algorithms that are built upon the inspirations of brain neuroscience. In this work, to further strengthen the bidirectional link between the 2 communities and especially promote the research on modern RL technology, we provide a comprehensive survey of recent advances in the area of brain-inspired/related RL algorithms. We start with basis theories of RL, and present a concise introduction to brain neuroscience related to RL. Then, we classify these advanced RL methodologies into 3 categories according to different connections of the brain, i.e., micro-neural activity, macro-brain structure, and cognitive function. Each category is further surveyed by presenting several modern RL algorithms along with their mathematical models, correlations with the brain, and open issues. Finally, we introduce several important applications of RL algorithms, followed by the discussions of challenges and opportunities for future research.

The transition metal trichalcogenides (TMTCs) are quasi-one-dimensional (1D) MX3-type van der Waals layered semiconductors, where M is a transition metal element of groups IV and V, and X indicates chalcogen element. Due to the unique quasi-1D crystalline structures, they possess several novel electrical properties such as variable bandgaps, charge density waves, and superconductivity, and highly anisotropic optical, thermoelectric, and magnetic properties. The study of TMTCs plays an essential role in the 1D quantum materials field, enabling new opportunities in the material research dimension. Currently, tremendous progress in both materials and solid-state devices has been made, demonstrating promising applications in the realization of nanoelectronic devices. This review provides a comprehensive overview to survey the state of the art in materials, devices, and applications based on TMTCs. Firstly, the symbolic structure, current primary synthesis methods, and physical properties of TMTCs have been discussed. Secondly, examples of TMTC applications in various fields are presented, such as photodetectors, energy storage devices, catalysts, and sensors. Finally, we give an overview of the opportunities and future perspectives for the research of TMTCs, as well as the challenges in both basic research and practical applications.

Heterogeneous catalysts, especially metal oxides, play a curial role in improving energy conversion efficiency and production of valuable chemicals. However, the surface structure at the atomic level and the nature of active sites are still ambiguous due to the dynamism of surface structure and difficulty in structure characterization under electrochemical conditions. This paper describes a strategy of the multiscale simulation to investigate the SnO_x reduction process and to build a structure–performance relation of SnO_x for CO₂ electroreduction. Employing high-dimensional neural network potential accelerated molecular dynamics and stochastic surface walking global optimization, coupled with density functional theory calculations, we propose that SnO₂ reduction is accompanied by surface reconstruction and charge density redistribution of active sites. A regulatory factor, the net charge, is identified to predict the adsorption capability for key intermediates on active sites. Systematic electronic analyses reveal the origin of the interaction between the adsorbates and the active sites. These findings uncover the quantitative correlation between electronic structure properties and the catalytic performance of SnO_x so that Sn sites with moderate charge could achieve the optimally catalytic performance of the CO₂ electroreduction to formate.

Osteoarthritis (OA) is always characterized as excessive reactive oxygen species (ROS) inside articular cavity. Mimicking natural metalloenzymes with metal ions as the active centers, stable metal organic framework (MOF) formed by natural polyphenols and metal ions shows great potential in alleviating inflammatory diseases. Herein, a series of novel copper-morin-based MOF (CuMHs) with different molar ratios of Cu²⁺ and MH were employed to serve as ROS scavengers for OA therapy. As a result, CuMHs exhibited enhanced dispersion in aqueous solution, improved biocompatibility, and efficient ROS-scavenging ability compared to MH. On the basis of H₂O₂-stimulated chondrocytes, intracellular ROS levels were efficiently declined and cell death was prevented after treated by Cu₆MH (Cu²⁺ and MH molar ratio of 6:1). Meanwhile, Cu₆MH also exhibited efficient antioxidant and anti-inflammation function by down-regulating the expression of IL6, MMP13, and MMP3, and up-regulating cartilage specific gene expression as well. Importantly, Cu₆MH could repair mitochondrial function by increasing mitochondrial membrane potential, reducing the accumulation of calcium ions, as well as promoting ATP content production. In OA joint model, intra-articular (IA) injected Cu₆MH suppressed the progression of OA. It endowed that Cu₆MH might be promising nanoenzymes for the prevention and treatment of various inflammatory diseases.

Protolignin extraction can facilitate structure elucidation and valorization of lignin in biorefinery, but is rather challenging due to the complex chemical bonds present. Here, we developed the in situ generated NH₃-reline (IGNR) system to realize one-pot protolignin extraction from lignocellulose. In the IGNR system, reline consisting of choline chloride and urea acted as both a solvent and a nucleophile generator, and the nucleophilic addition–elimination mechanism was verified by model compound studies. The in situ generated NH₃ could precisely cleave the lignin–carbohydrate esters in lignocellulose with a near-quantitative retention of carbohydrates. The extracted IGNR–Protolignin exhibited native lignin substructure with high molecular weight and high β-O-4′ content (41.5 per 100 aromatic units). In addition, the up-scaled kilogram reaction demonstrated the feasibility of the IGNR system for potential industrial application in a green and sustainable pathway. This work represents a breakthrough toward protolignin extraction in practice with the future goal of achieving total biorefinery.

Platinum-based chemotherapy remains the main systemic treatment of ovarian cancer (OC). However, the inevitable development of platinum and poly (adenosine diphosphate-ribose) polymerase inhibitor (PARPi) resistance is associated with poor outcomes, which becomes a major obstacle in the management of this disease. The present study developed “all-in-one” nanoparticles that contained the PARPi olaparib and gallium (Ga) (III) (olaparib-Ga) to effectively reverse PARPi resistance in platinum-resistant A2780-cis and SKOV3-cis OC cells and in SKOV3-cis tumor models. Notably, the olaparib-Ga suppressed SKOV3-cis tumor growth with negligible toxicity. Moreover, the suppression effect was more evident when combining olaparib-Ga with cisplatin or carboplatin, as evaluated in A2780-cis and SKOV3-cis cells. Mechanistically, the combined treatment induced DNA damage, which elicited the activation of ataxia telangiectasia mutated (ATM)/AMT- and Rad3-related (ATR) checkpoint kinase 1 (Chk1)/Chk2 signal transduction pathways. This led to the arrest of cell cycle progression at S and G₂/M phases, which eventually resulted in apoptosis and cell death due to unrepairable DNA damage. In addition, effective therapeutic responses to olaparib-Ga and cisplatin combination or olaparib-Ga and carboplatin combination were observed in SKOV3-cis tumor-bearing animal models. Altogether, the present findings demonstrate that olaparib-Ga has therapeutic implications in platinum-resistant OC cells, and the combination of olaparib-Ga with cisplatin or carboplatin may be promising for treating patients with OC who exhibit resistance to both PARPi and platinum.

This work aims to explore the impact of Digital Twins Technology on industrial manufacturing in the context of Industry 5.0. A computer is used to search the Web of Science database to summarize the Digital Twins in Industry 5.0. First, the background and system architecture of Industry 5.0 are introduced. Then, the potential applications and key modeling technologies in Industry 5.0 are discussd. It is found that equipment is the infrastructure of industrial scenarios, and the embedded intelligent upgrade for equipment is a Digital Twins primary condition. At the same time, Digital Twins can provide automated real-time process analysis between connected machines and data sources, speeding up error detection and correction. In addition, Digital Twins can bring obvious efficiency improvements and cost reductions to industrial manufacturing. Digital Twins reflects its potential application value and subsequent potential value in Industry 5.0 through the prospect. It is hoped that this relatively systematic overview can provide technical reference for the intelligent development of industrial manufacturing and the improvement of the efficiency of the entire business process in the Industrial X.0 era.

We study the mechanism of topological mass generation for 3-dimensional Chern–Simons gauge theories and propose a brand-new topological equivalence theorem to connect scattering amplitudes of the physical gauge boson states to that of the transverse states under high-energy expansion. We prove a general energy cancelation mechanism for N-point physical gauge boson amplitudes, which predicts large cancelations of E^{4 − L} → E^{(4 − L) − N} at any L-loop level (L ⩾ 0). We extend the double-copy approach to construct massive graviton amplitudes and to study their structures. We newly uncovered a series of strikingly large energy cancelations E¹² → E¹ of the tree-level 4-graviton scattering amplitude under high-energy expansion and establish a new correspondence between the 2 energy cancelations in the topologically massive Yang–Mills gauge theory and the topologically massive gravity theory. We further study the scattering amplitudes of Chern–Simons gauge bosons and gravitons in the nonrelativistic limit.

Semiconductor quantum dots have been emerging as one of the most ideal materials for artificial photosynthesis. Here, we report the assembled ZnS-CdS hybrid heterostructure for efficient coupling cooperative redox catalysis toward the oxidation of 1-phenylethanol to acetophenone/2,3-diphenyl-2,3-butanediol (pinacol) integrated with the reduction of protons to H₂. The strong interaction and typical type-I band-position alignment between CdS quantum dots and ZnS quantum dots result in efficient separation and transfer of electron-hole pairs, thus distinctly enhancing the coupled photocatalyzed-redox activity and stability. The optimal ZnS-CdS hybrid also delivers a superior performance for various aromatic alcohol coupling photoredox reaction, and the ratio of electrons and holes consumed in such redox reaction is close to 1.0, indicating a high atom economy of cooperative coupling catalysis. In addition, by recycling the scattered light in the near field of a SiO₂ sphere, the SiO₂-supported ZnS-CdS (denoted as ZnS-CdS/SiO₂) catalyst can further achieve a 3.5-fold higher yield than ZnS-CdS hybrid. Mechanistic research clarifies that the oxidation of 1-phenylethanol proceeds through the pivotal radical intermediates of ^•C(CH₃)(OH)Ph. This work is expected to promote the rational design of semiconductor quantum dots-based heterostructured catalysts for coupling photoredox catalysis in organic synthesis and clean fuels production.

The power generated by flexible wearable devices (FWDs) is normally insufficient to eradicate bacteria, and many conventional antibacterial strategies are also not suitable for flexible and wearable applications because of the strict mechanical and electrical requirements. Here, polypyrrole (PPy), a conductive polymer with a high mass density, is used to form a nanostructured surface on FWDs for antibacterial purposes. The conductive films with PPy nanorods (PNRs) are found to sterilize 98.2 ± 1.6% of Staphylococcus aureus and 99.6 ± 0.2% of Escherichia coli upon mild electrification (1 V). Bacteria killing stems from membrane stress produced by the PNRs and membrane depolarization caused by electrical neutralization. Additionally, the PNR films exhibit excellent biosafety and electrical stability. The results represent pioneering work in fabricating antibacterial components for FWDs by comprehensively taking into consideration the required conductivity, mechanical properties, and biosafety.

Monitoring the physiological changes of organelles is essential for understanding the local biological information of cells and for improving the diagnosis and therapy of diseases. Currently, fluorescent probes are considered as the most powerful tools for imaging and have been widely applied in biomedical fields. However, the expected targeting effects of these probes are often inconsistent with the real experiments. The design of fluorescent probes mainly depends on the empirical knowledge of researchers, which was inhibited by limited chemical space and low efficiency. Herein, we proposed a novel multilevel framework for the prediction of organelle-targeted fluorescent probes by employing advanced artificial intelligence algorithms. In this way, not only the targeting mechanism could be interpreted beyond intuitions but also a quick evaluation method could be established for the rational design. Furthermore, the targeting and imaging powers of the optimized and synthesized probes based on this methodology were verified by quantitative calculation and experiments.

Overexpression of CD47 is frequently observed in various types of human malignancies, inhibiting myeloid-mediated elimination of tumor cells and affecting the prognosis of cancer patients. By mapping biomarker expression, immuno-positron emission tomography has been increasingly used for patient screening and response monitoring. By immunization alpacas with recombinant human CD47, we prepared a CD47-targeting nanobody C2 and developed [⁶⁸Ga]Ga-NOTA-C2, followed by an exploration of the diagnostic value in CD47-expressing tumor models including gastric-cancer patient-derived xenograft models. By fusing C2 to an albumin binding domain (ABD), we synthesized ABDC2, which had increased in vivo half-life and improved targeting properties. We further labeled ABDC2 with ⁶⁸Ga/⁸⁹Zr/¹⁷⁷Lu to develop radionuclide theranostic pairs and evaluated the pharmacokinetics and theranostic efficacies of the agents in cell- and patient-derived models. Both C2 and ABDC2 specifically reacted with human CD47 with a high K_D value of 23.50 and 84.57 pM, respectively. [⁶⁸Ga]Ga-NOTA-C2 was developed with high radiochemical purity (99 >%, n = 4) and visualized CD47 expression in the tumors. In comparison to the rapid renal clearance and short half-life of [⁶⁸Ga]Ga-NOTA-C2, both [⁶⁸Ga]Ga-NOTA-ABDC2 and [⁸⁹Zr]Zr-DFO-ABDC2 showed prolonged circulation and increased tumor uptake, with the highest uptake of [⁸⁹Zr]Zr-DFO-ABDC2 occurring at 72 h post-injection. Moreover, [¹⁷⁷Lu]Lu-DOTA-ABDC2 radioimmunotherapy suppressed the tumor growth but was associated with toxicity, warranting further optimization of the treatment schedules. Taken together, we reported a series of nanobody-derived CD47-targeted agents, of which [⁶⁸Ga]Ga-NOTA-C2 and [⁸⁹Zr]Zr-DFO-ABDC2 are readily translatable. Optimization and translation of CD47-targeted theranostic pair may provide new prospects for CD47-targeted management of solid tumors.

To elucidate the role of post-translational modifications (PTMs) in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein’s structure and virulence, we generated a high-resolution map of 87 PTMs using liquid chromatography with tandem mass spectrometry data on the extracted spike protein from SARS-CoV-2 virions and then reconstituted its structure heterogeneity caused by PTMs. Nonetheless, Alphafold2, a high-accuracy artificial intelligence tool to perform protein structure prediction, relies solely on primary amino acid sequence, whereas the impact of PTM, which often modulates critical protein structure and function, is much ignored. To overcome this challenge, we proposed the mutagenesis approach—an in silico, site-directed amino acid substitution to mimic the influence of PTMs on protein structure due to altered physicochemical properties in the post-translationally modified amino acids—and then reconstituted the spike protein’s structure from the substituted sequences by Alphafold2. For the first time, the proposed method revealed predicted protein structures resulting from PTMs, a problem that Alphafold2 has yet to address. As an example, we performed computational analyses of the interaction of the post-translationally modified spike protein with its host factors such as angiotensin-converting enzyme 2 to illuminate binding affinity. Mechanistically, this study suggested the structural analysis of post-translationally modified protein via mutagenesis and deep learning. To summarize, the reconstructed spike protein structures showed that specific PTMs can be used to modulate host factor binding, guide antibody design, and pave the way for new therapeutic targets. The code and Supplementary Materials are freely available at https://github.com/LTZHKUSTGZ/SARS-CoV-2-spike-protein-PTM.

Transition metal-based single-atom catalysts (TM-SACs) are promising alternatives to Au- and Ag-based electrocatalysts for CO production through CO₂ reduction reaction. However, developing TM-SACs with high activity and selectivity at low overpotentials is challenging. Herein, a novel Fe-based SAC with Si doping (Fe-N-C-Si) was prepared, which shows a record-high electrocatalytic performance toward the CO₂-to-CO conversion with exceptional current density (>350.0 mA cm⁻²) and ~100% Faradaic efficiency (FE) at the overpotential of <400 mV, far superior to the reported Fe-based SACs. Further assembling Fe-N-C-Si as the cathode in a rechargeable Zn-CO₂ battery delivers an outstanding performance with a maximal power density of 2.44 mW cm⁻² at an output voltage of 0.30 V, as well as high cycling stability and FE (>90%) for CO production. Experimental combined with theoretical analysis unraveled that the nearby Si dopants in the form of Si-C/N bonds modulate the electronic structure of the atomic Fe sites in Fe-N-C-Si to markedly accelerate the key pathway involving *CO intermediate desorption, inhibiting the poisoning of the Fe sites under high CO coverage and thus boosting the CO₂RR performance. This work provides an efficient strategy to tune the adsorption/desorption behaviors of intermediates on single-atom sites to improve their electrocatalytic performance.

F-actin (filamentous actin) has been shown to be sensitive to mechanical stimuli and play critical roles in cell attachment, migration, and cancer metastasis, but there are very limited ways to perturb F-actin dynamics with low cell toxicity. Magnetic field is a noninvasive and reversible physical tool that can easily penetrate cells and human bodies. Here, we show that 0.1/0.4-T 4.2-Hz moderate-intensity low-frequency rotating magnetic field-induced electric field could directly decrease F-actin formation in vitro and in vivo, which results in decreased breast cancer cell migration, invasion, and attachment. Moreover, low-frequency rotating magnetic fields generated significantly different effects on F-actin in breast cancer vs. noncancerous cells, including F-actin number and their recovery after magnetic field retrieval. Using an intermittent treatment modality, low-frequency rotating magnetic fields could significantly reduce mouse breast cancer metastasis, prolong mouse survival by 31.5 to 46.0% (P < 0.0001), and improve their overall physical condition. Therefore, our work demonstrates that low-frequency rotating magnetic fields not only can be used as a research tool to perturb F-actin but also can inhibit breast cancer metastasis through F-actin modulation while having minimum effects on normal cells, which reveals their potential to be developed as temporal-controlled, noninvasive, and high-penetration physical treatments for metastatic cancer.

Electroactive biofilm plays a crucial rule in the electron transfer efficiency of microbial electrochemical systems (MES). However, the low ability to form biofilm and the low conductivity of the formed biofilm substantially limit the extracellular electron transfer rate of microbial cells to the electrode surfaces in MES. To promote biofilm formation and enhance biofilm conductivity, we develop synthetic biology approach to systematically engineer Shewanella oneidensis, a model exoelectrogen, via modular manipulation of the full-cycle different stages of biofilm formation, namely, from initial contact, cell adhesion, and biofilm growth stable maturity to cell dispersion. Consequently, the maximum output power density of the engineered biofilm reaches 3.62 ± 0.06 W m⁻², 39.3-fold higher than that of the wild-type strain of S. oneidensis, which, to the best our knowledge, is the highest output power density that has ever been reported for the biofilms of the genetically engineered Shewanella strains.

The discovery and study of skyrmion materials play an important role in basic frontier physics research and future information technology. The database of 196 materials, including 64 skyrmions, was established and predicted based on machine learning. A variety of intrinsic features are classified to optimize the model, and more than a dozen methods had been used to estimate the existence of skyrmion in magnetic materials, such as support vector machines, k-nearest neighbor, and ensembles of trees. It is found that magnetic materials can be more accurately divided into skyrmion and non-skyrmion classes by using the classification of electronic layer. Note that the rare earths are the key elements affecting the production of skyrmion. The accuracy and reliability of random undersampling bagged trees were 87.5% and 0.89, respectively, which have the potential to build a reliable machine learning model from small data. The existence of skyrmions in LaBaMnO is predicted by the trained model and verified by micromagnetic theory and experiments.

Drug therapy is among the most widely used methods in disease treatment. However, there remains a trade-off problem between drug dosage and toxicity. Blood purification by adsorption of excessive drugs during clinical treatment could be a solution for enhancing therapeutic efficacy while maintaining normal body function. Here, inspired by the intrinsic action mechanism of chemotherapeutic agents in targeting DNA in the cell nucleus, we present DNA–polyelectrolyte composite responsive microparticles for chemotherapeutics cleaning. The presence of DNA in the microparticles enabled the adsorption of multiple common chemotherapy drugs. Moreover, the microparticles are endowed with a porous structure and a photothermal-responsive ability, both of which contribute to improved adsorption by enhancing the contact of the microparticles with the drug solution. On the basis of that, the microparticles are integrated into a herringbone-structured microfluidic chip. The fluid mixing capacity and the enhanced drug cleaning efficiency of the microfluidic platform are validated on-chip. These results indicate the value of the DNA–polyelectrolyte composite responsive microparticles for drug capture and blood purification. We believe the microparticle-integrated microfluidic platform could provide a solution for settling the dosage–toxicity trade-off problems in chemotherapy.

Diverse defects in copper indium gallium diselenide solar cells cause nonradiative recombination losses and impair device performance. Here, an organic passivation scheme for surface and grain boundary defects is reported, which employs an organic passivation agent to infiltrate the copper indium gallium diselenide thin films. A transparent conductive passivating (TCP) film is then developed by incorporating metal nanowires into the organic polymer and used in solar cells. The TCP films have a transmittance of more than 90% in the visible and nearinfrared spectra and a sheet resistance of ~10.5 Ω/sq. This leads to improvements in the open-circuit voltage and the efficiency of the organic passivated solar cells compared with control cells and paves the way for novel approaches to copper indium gallium diselenide defect passivation and possibly other compound solar cells.

Integrating mechanical computing functions into robotic materials, microelectromechanical systems, or soft robotics can improve their intelligence in stimulation–response processes. Current mechanical computing systems exhibit limitations, including incomplete functions, unchangeable computing rules, difficulties in realizing random logic, and lack of reusability. To overcome these limitations, we propose a straightforward method of designing mechanical computing systems—based on the logic expressions—for complex computations. We designed soft, B-shaped mechanical metamaterial units, and compressed them to render stress inputs; the outputs are represented by the light-shielding effects caused by the unit deformations. We realized logic gates and corresponding combinations (including half/full binary adder/subtractor and addition/subtraction of 2 numbers with multiple bits) and provided a versatile solution for making a mechanical analog-to-digital converter to generate both ordered and disordered numbers. We performed all of the computations within the elastic regions of the B-shaped units; thus, after one computation, the systems can return to the initial states for reuse. The proposed mechanical computers will potentially enable robotic materials, microelectromechanical systems, or soft robotics to perform complex tasks. Furthermore, one can extend this concept to systems that are based on other mechanisms or materials.

The study of lipid metabolism relies on the characterization of the lipidome, which is quite complex due to the structure variations of the lipid species. New analytical tools have been developed recently for characterizing fine structures of lipids, with C=C location identification as one of the major improvements. In this study, we studied the lipid metabolism reprograming by analyzing glycerol phospholipid compositions in breast cancer cell lines with structural specification extended to the C=C location level. Inhibition of the lipid desaturase, stearoyl-CoA desaturase 1, increased the proportion of n-10 isomers that are produced via an alternative fatty acid desaturase 2 pathway. However, there were different variations of the ratio of n-9/n-7 isomers in C18:1-containing glycerol phospholipids after stearoyl-CoA desaturase 1 inhibition, showing increased tendency in MCF-7 cells, MDA-MB-468 cells, and BT-474 cells, but decreased tendency in MDA-MB-231 cells. No consistent change of the ratio of n-9/n-7 isomers was observed in SK-BR-3 cells. This type of heterogeneity in reprogrammed lipid metabolism can be rationalized by considering both lipid desaturation and fatty acid oxidation, highlighting the critical roles of comprehensive lipid analysis in both fundamental and biomedical applications.

Drug delivery systems with high-targeted doses can minimize excipients, reduce side effects, and improve efficacy. Human blood circulation is a complex circulatory system, and the motion control of microrobots in the static flow field in vitro is completely different from in vivo. How to achieve precise counterflow motion for targeted drug delivery without vascular blockage and immune rejection is the biggest challenge for micro-nano robots. Here, we propose a control method that enables vortex-like paramagnetic nanoparticle swarm (VPNS) to move upstream against the flow. By mimicking the clustering motion of wild herring schools and the rolling of leukocytes, VPNS are incredibly stable even when subjected to high-intensity jet impacts in the blood environment, can travel upstream, anchor at the target location, and dissipate when the magnetic field is withdrawn, which greatly reduces the risk of thrombosis. VPNS can also upstream along the vessel wall without an additional energy source and has a marked targeted therapeutic effect on subcutaneous tumors.

Rapid and effective repair of injured or diseased bone defects remains a major challenge due to shortages of implants. Smart hydrogels that respond to internal and external stimuli to achieve therapeutic actions in a spatially and temporally controlled manner have recently attracted much attention for bone therapy and regeneration. These hydrogels can be modified by introducing responsive moieties or embedding nanoparticles to increase their capacity for bone repair. Under specific stimuli, smart hydrogels can achieve variable, programmable, and controllable changes on demand to modulate the microenvironment for promoting bone healing. In this review, we highlight the advantages of smart hydrogels and summarize their materials, gelation methods, and properties. Then, we overview the recent advances in developing hydrogels that respond to biochemical signals, electromagnetic energy, and physical stimuli, including single, dual, and multiple types of stimuli, to enable physiological and pathological bone repair by modulating the microenvironment. Then, we discuss the current challenges and future perspectives regarding the clinical translation of smart hydrogels.

Organic scintillators, materials with the ability to exhibit luminescence when exposed to X-rays, have aroused increasing interest in recent years. However, the enhancement of radioluminescence and improving X-ray absorption of organic scintillators lie in the inherent dilemma, due to the waste of triplet excitons and weak X-ray absorption during scintillation. Here, we employ halogenated thermally activated delayed fluorescence materials to improve the triplet exciton utilization and X-ray absorption simultaneously, generating efficient scintillation with a low detection limit, which is one order of magnitude lower than the dosage for X-ray medical diagnostics. Through experimental study and theoretical calculation, we reveal the positive role of X-ray absorption, quantum yields of prompt fluorescence, and intersystem crossing in promoting the radioluminescence intensity. This finding offers an opportunity to design diverse types of organic scintillators and expands the applications of thermally activated delayed fluorescence.

Borrowing from natural mechanisms for material design can lead to functional mimicry and improvement. Inspired by graphite formation, a thermopressure coupling strategy under micropressure (<400 Pa) is applied to prepare carbon anodes. A thermopressure response is discovered based on the cellulose precursor. Here, homologous graphene quantum dot/hard carbon (GQD/HC) heterostructures are synthesized. Under 181.4 Pa and 1,200 °C, the product shows a capacity of 310 mAh g⁻¹, while the capacity of the direct carbonization product is only 120 mAh g⁻¹. Prominently, the GQD/HC heterostructure displays marked mechanical strength and flexibility. The experimental and theoretical results illustrate the ion and electron transfer, coordination environment, and electronic states in the GQD/HC heterostructure and elaborate on the origin of the enhanced performance. The thermopressure coupling under micropressure mimics graphite formation, but the heterostructure has better properties than traditional carbon materials. Additionally, micropressure injects new vitality into material research.

HfO₂-based ferroelectrics have evoked considerable interest owing to the complementary metal–oxide semiconductor compatibility and robust ferroelectricity down to a few unit cells. However, the unique wake-up effect of HfO₂-based ferroelectric films severely restricts the improvement of their performance. In particular, the domain structure is an important characteristic of ferroelectric materials, which still has not been well understood in HfO₂-based ferroelectrics. In this work, a Hf_0.5Zr_0.5O₂ ferroelectric thin film is grown on a typical Si substrate buffered with TiN electrode. The 90° domains of the Pca2₁ ferroelectric phase with head-to-tail and tail-to-tail structures can be observed by Cs-corrected scanning transmission electron microscope under their pristine condition. After waking up, the 180° domain is displayed in the ferroelectric phase. The remarkable differences in domain walls for 90° and 180° domains are characterized by qualitatively mapping the polarization distributions at the atomic scale. The domain wall changes from the \left[10\overline{1}\right] of the Hf_0.5Zr_0.5O₂ film to the [001] of the Hf_0.5Zr_0.5O₂ film. This result provides fundamental information for understanding the domain structure of HfO₂-based ferroelectrics.

Intelligent stimuli-responsive fluorescence materials are extremely pivotal for fabricating luminescent turn-on switching in solid-state photonic integration technology, but it remains a challenging objective for typical 3-dimensional (3D) perovskite nanocrystals. Herein, by fine-tuning the accumulation modes of metal halide components to dynamically control the carrier characteristics, a novel triple-mode photoluminescence (PL) switching was realized in 0D metal halide through stepwise single-crystal to single-crystal (SC-SC) transformation. Specifically, a family of 0D hybrid antimony halides was designed to exhibit three distinct types of PL performance including nonluminescent [Ph₃EtP]₂Sb₂Cl₈ (1), yellow-emissive [Ph₃EtP]₂SbCl₅·EtOH (2), and red-emissive [Ph₃EtP]₂SbCl₅ (3). Upon stimulus of ethanol, 1 was successfully converted to 2 through SC-SC transformation with enhanced PL quantum yield from ~0% to 91.50% acting as “turn-on” luminescent switching. Meanwhile, reversible SC-SC and luminescence transformation between 2 and 3 can be also achieved in the ethanol impregnation–heating process as luminescence vapochromism switching. As a consequence, a new triple-model turn-on and color-adjustable luminescent switching of off–on^I–on^II was realized in 0D hybrid halides. Simultaneously, wide advanced applications were also achieved in anti-counterfeiting, information security, and optical logic gates. This novel photon engineering strategy is expected to deepen the understanding of dynamic PL switching mechanism and guide development of new smart luminescence materials in cutting-edge optical switchable device.

The enzyme-mimicking catalytic activity of single-atom nanozymes has been widely used in tumor treatment. However, research on alleviating metabolic diseases, such as hyperglycemia, has not been reported. Herein, we found that the single-atom Ce-N₄-C-(OH)₂ (SACe-N₄-C-(OH)₂) nanozyme promoted glucose absorption in lysosomes, resulting in increased reactive oxygen species production in HepG2 cells. Furthermore, the SACe-N₄-C-(OH)₂ nanozyme initiated a cascade reaction involving superoxide dismutase-, oxidase-, catalase-, and peroxidase-like activity to overcome the limitations associated with the substrate and produce •OH, thus improving glucose intolerance and insulin resistance by increasing the phosphorylation of protein kinase B and glycogen synthase kinase 3β, and the expression of glycogen synthase, promoting glycogen synthesis to improve glucose intolerance and insulin resistance in high-fat diet-induced hyperglycemic mice. Altogether, these results demonstrated that the novel nanozyme SACe-N₄-C-(OH)₂ alleviated the effects of hyperglycemia without evident toxicity, demonstrating its excellent clinical application potential.

While conventional nanosystems can target infected lung tissue, they cannot achieve precise cellular targeting and enhanced therapy by modulating inflammation and microbiota for effective therapy. Here, we designed a nucleus-targeted nanosystem with adenosine triphosphate (ATP) and reactive oxygen species stimuli–response to treat pneumonia coinfected with bacteria and virus that is enhanced through inflammation and microbiota regulation. The nucleus-targeted biomimetic nanosystem was prepared through the combined bacteria–macrophage membrane and loaded hypericin and ATP-responsive dibenzyl oxalate (MMHP) subsequently. The MMHP despoiled the Mg²⁺ of intracellular cytoplasm in bacteria to achieve an effective bactericidal performance. Meanwhile, MMHP can target the cell nucleus and inhibit the H1N1 virus duplication by inhibiting the activity of nucleoprotein. MMHP possessed an immunomodulatory ability to reduce the inflammatory response and activate CD8⁺ T cells for assisted infection elimination. During the mice model, the MMHP effectively treated pneumonia coinfected with Staphylococcus aureus and H1N1 virus. Meanwhile, MMHP mediated the composition of gut microbiota to enhance the pneumonia therapy. Therefore, the dual stimuli-responsive MMHP possessed promising clinical translational potential to therapy infectious pneumonia.

It is well known that alcohol consumption leads to substantially increased free radical levels and health risks, which lacks effective treatment besides alcohol abstinence. Here, we compared different static magnetic field (SMF) settings and found that a downward quasi-uniform SMF of ~0.1 to 0.2 T could effectively alleviate alcohol-induced liver damage and lipid accumulation and improve hepatic function. SMFs of two different directions can reduce the inflammation, reactive oxygen species levels, and oxidative stress in the liver, while the downward SMF had more obvious effects. Moreover, we found that the upward direction SMF of ~0.1 to 0.2 T could inhibit DNA synthesis and regeneration in hepatocytes, which caused detrimental effects on the lifespan of "heavy drinking" mice. In contrast, the downward SMF prolongs survival of "heavy drinking" mice. On one hand, our study shows that ~0.1 to 0.2 T moderate quasi-uniform SMFs with a downward direction have great promises to be developed into a physical method to reduce alcohol-induced liver damage; on the other hand, although the internationally recognized upper limit for SMF public exposure is 0.4 T, people should also pay extra attention to SMF strength, direction, and inhomogeneity that could generate harmful effects on specific severe pathological conditions.

Traditional metal oxide semiconductor (MOS) gas sensors have limited applications in wearable devices owing to their inflexibility and high-power consumption by substantial heat loss. To overcome these limitations, we prepared doped Si/SiO₂ flexible fibers by a thermal drawing method as substrates to fabricate MOS gas sensors. A methane (CH₄) gas sensor was demonstrated by subsequently in situ synthesizing Co-doped ZnO nanorods on the fiber surface. The doped Si core acted as the heating source through Joule heating, which conducted heat to the sensing material with reduced heat loss; the SiO₂ cladding was an insulating substrate. The gas sensor was integrated into a miner cloth as a wearable device, and the concentration change of CH₄ was monitored in real time through different colored light-emitting diodes. Our study demonstrated the feasibility of using doped Si/SiO₂ fibers as the substrates to fabricate wearable MOS gas sensors, where the sensors have substantial advantages over tradition sensors in flexibility, heat utilization, etc.

The stimulator of interferon genes (STING) protein is an important and promising innate immune target for tumor therapy. However, the instability of the agonists of STING and their tendency to cause systemic immune activation is a hurdle. The STING activator, cyclic di-adenosine monophosphate (CDA), produced by the modified Escherichia coli Nissle 1917, shows high antitumor activity and effectively reduces the systemic effects of the “off-target” caused by the activation of the STING pathway. In this study, we used synthetic biological approaches to optimize the translation levels of the diadenylate cyclase that catalyzes CDA synthesis in vitro. We developed 2 engineered strains, CIBT4523 and CIBT4712, for producing high levels of CDA while keeping their concentrations within a range that did not compromise the growth. Although CIBT4712 exhibited stronger induction of the STING pathway corresponding to in vitro CDA levels, it had lower antitumor activity than CIBT4523 in an allograft tumor model, which might be related to the stability of the surviving bacteria in the tumor tissue. CIBT4523 exhibited complete tumor regression, prolonged survival of mice, and rejection of rechallenged tumors, thus, offering new possibilities for more effective tumor therapy. We showed that the appropriate production of CDA in engineered bacterial strains is essential for balancing antitumor efficacy and self-toxicity.