Search

The communication between the 5’ 7-methylguanosine cap structure and the 3’ poly(A) tail of eukaryotic mRNAs plays a crucial role in translation regulation. Scientists hypothesized that mRNAs interact with the cap-binding protein eIF4E, the poly(A)-binding protein PABP1, and eIF4GI through interactions with the 5’ cap structure and the 3’ poly(A) tail. These interactions bring the 5’ end and the 3’ end of the mRNA into proximity known as the eukaryotic “closed-loop” model. The formation of this “closed-loop” structure allows the physical communication between the 5’ end and the 3’ end of the mRNA, and it is thought to enhance translation initiation. Although the “closed-loop” model is a widely accepted hypothesis, there is still conflicting evidence both supporting and opposing this model.To further elucidate this model, we tested the formation of the “pseudo-closed-loop” structure with human proteins in vitro using the electrophoretic mobility shift assay (EMSA). However, instead of forming the single “pseudo-closed-loop” structure, our results show that two non-interacting complexes are formed in vitro: the 5’cap•eIF4E•eIF4GI complex and the poly(A)-tail•PABP1•eIF4GI complex. The formation of these two separate complexes suggests that the binding of eIF4GI to eIF4E and PABP1 is competitive and mutually exclusive. Overall, we hypothesize that the formation of the “pseudo-closed-loop” structure may not occur in all mRNA transcripts or all contexts of eukaryotic translation initiation.

In recent years, the growing pervasiveness of wearable technology has created new opportunities for medical and emergency rescue operations to protect users' health and safety, such as cost-effective medical solutions, more convenient healthcare and quick hospital treatments, which make it easier for the Internet of Medical Things (IoMT) to evolve. The study first presents an overview of the IoMT before introducing the IoMT architecture. Later, it portrays an overview of the core technologies of the IoMT, including cloud computing, big data and artificial intelligence, and it elucidates their utilization within the healthcare system. Further, several emerging challenges, such as cost-effectiveness, security, privacy, accuracy and power consumption, are discussed, and potential solutions for these challenges are also suggested.

In recent years, the growing pervasiveness of wearable technology has created new opportunities for medical and emergency rescue operations to protect users' health and safety, such as cost-effective medical solutions, more convenient healthcare and quick hospital treatments, which make it easier for the Internet of Medical Things (IoMT) to evolve. The study first presents an overview of the IoMT before introducing the IoMT architecture. Later, it portrays an overview of the core technologies of the IoMT, including cloud computing, big data and artificial intelligence, and it elucidates their utilization within the healthcare system. Further, several emerging challenges, such as cost-effectiveness, security, privacy, accuracy and power consumption, are discussed, and potential solutions for these challenges are also suggested.

In recent years, the growing pervasiveness of wearable technology has created new opportunities for medical and emergency rescue operations to protect users' health and safety, such as cost-effective medical solutions, more convenient healthcare and quick hospital treatments, which make it easier for the Internet of Medical Things (IoMT) to evolve. The study first presents an overview of the IoMT before introducing the IoMT architecture. Later, it portrays an overview of the core technologies of the IoMT, including cloud computing, big data and artificial intelligence, and it elucidates their utilization within the healthcare system. Further, several emerging challenges, such as cost-effectiveness, security, privacy, accuracy and power consumption, are discussed, and potential solutions for these challenges are also suggested.

Tsunamigenic earthquakes pose considerable risks, both economically and socially, yet earthquake and tsunami hazard assessments are typically conducted separately. Earthquakes associated with unexpected tsunamis, such as the 2018 Mw 7.5 strike-slip Sulawesi earthquake, emphasize the need to study the tsunami potential of active submarine faults in different tectonic settings. Here, we investigate physics-based scenarios combining simulations of 3D earthquake dynamic rupture and seismic wave propagation with tsunami generation and propagation. We present time-dependent modeling of one-way linked and 3D fully coupled earthquakes and tsunamis for the ∼ 100 km long Húsavík–Flatey Fault Zone (HFFZ) in North Iceland. Our analysis shows that the HFFZ has the potential to generate sizable tsunamis. The six dynamic rupture models sourcing our tsunami scenarios vary regarding hypocenter location, spatiotemporal evolution, fault slip, and fault structure complexity but coincide with historical earthquake magnitudes. Earthquake dynamic rupture scenarios on a less segmented fault system, particularly with a hypocenter location in the eastern part of the fault system, have a larger potential for local tsunami generation. Here, dynamically evolving large shallow fault slip (∼ 8 m), near-surface rake rotation (± 20∘), and significant coseismic vertical displacements of the local bathymetry (± 1 m) facilitate strike-slip faulting tsunami generation. We model tsunami crest to trough differences (total wave heights) of up to ∼ 0.9 m near the town Ólafsfjörður. In contrast, none of our scenarios endanger the town of Akureyri, which is shielded by multiple reflections within the narrow Eyjafjörður bay and by Hrísey island. We compare the modeled one-way linked tsunami waveforms with simulation results using a 3D fully coupled approach. We find good agreement in the tsunami arrival times and location of maximum tsunami heights. While seismic waves result in transient motions of the sea, This work received funding from the European Union's Horizon 2020 research and innovation programme (TEAR ERC Starting; grant no. 852992) and Horizon Europe (ChEESE-2P, grant no. 101093038; DT-GEO, grant no. 101058129; and Geo-INQUIRE, grant no. 101058518). We acknowledge additional funding from the National Science Foundation (grant no. EAR-2121666) and the National Aeronautics and Space Administration (grant no. 80NSSC20K0495)., Peer Reviewed, Postprint (published version)

In recent years, the growing pervasiveness of wearable technology has created new opportunities for medical and emergency rescue operations to protect users' health and safety, such as cost-effective medical solutions, more convenient healthcare and quick hospital treatments, which make it easier for the Internet of Medical Things (IoMT) to evolve. The study first presents an overview of the IoMT before introducing the IoMT architecture. Later, it portrays an overview of the core technologies of the IoMT, including cloud computing, big data and artificial intelligence, and it elucidates their utilization within the healthcare system. Further, several emerging challenges, such as cost-effectiveness, security, privacy, accuracy and power consumption, are discussed, and potential solutions for these challenges are also suggested.

Biao Hou,1– 3,* Bo Li,4,* Wanjun Deng,3,* Bo Li,1,2 Bibo Ren,5 Chao Hu,3 Guowei Zhang,1,2 Fen Yang,6 Meimei Xiao,3 Songlin Xie,3 Denghui Xie1,2 1Department of Joint Surgery and Sports Medicine, Center for Orthopedic Surgery, Orthopedic Hospital of Guangdong Province, The Third Affiliated Hospital of Southern Medical University, Guangzhou, People’s Republic of China; 2Guangdong Provincial Key Laboratory of Bone and Joint Degeneration Diseases, Academy of Orthopedics, Guangzhou, Guangdong Province, People’s Republic of China; 3Department of Hand and Foot Microsurgery, The affiliated Nanhua Hospital, Hengyang Medical College, University of South China, Hengyang, People’s Republic of China; 4Department of Radiology, West China Hospital, Sichuan University, Chengdu, People’s Republic of China; 5College of Biomass Science and Engineering, Sichuan University, Chengdu, People’s Republic of China; 6Department of Infectious Diseases, The Affiliated Nanhua Hospital, Hengyang Medical College, University of South China, Hengyang, Hunan, People’s Republic of China*These authors contributed equally to this workCorrespondence: Denghui Xie; Songlin Xie, Email smuspine@163.com; xiesonglin0929@163.comPurpose: This research was to innovate a nanozyme-based therapeutic strategy that combines aggregation-induced emission (AIE) photosensitizers with copper nanozymes. This approach is designed to address the hypoxic conditions often found in bacterial infections and aims to boost the effectiveness of photodynamic therapy (PDT) by ensuring sufficient oxygen supply for reactive oxygen species (ROS) generation.Methods: Our approach involved the synthesis of dihydroxyl triphenyl vinyl pyridine (DHTPY)-Cu@zoledronic acid (ZOL) nanozyme particles. We initially synthesized DHTPY and then combined it with copper nanozymes to form the DHTPY-Cu@ZOL composite. The nanozyme’s size, morphology, and chemical properties were characterized using various tech

In the context of epidemic spreading, many intricate dynamical patterns can emerge due to the cooperation of different types of pathogens or the interaction between the disease spread and other failure propagation mechanism. To unravel such patterns, simulation frameworks are usually adopted, but they are computationally demanding on big networks and subject to large statistical uncertainty. Here, we study the two-layer spreading processes on unidirectionally dependent networks, where the spreading infection of diseases or malware in one layer can trigger cascading failures in another layer and lead to secondary disasters, e.g., disrupting public services, supply chains, or power distribution. We utilize a dynamic message-passing method to devise efficient algorithms for inferring the system states, which allows one to investigate systematically the nature of complex intertwined spreading processes and evaluate their impact. Based on such dynamic message-passing framework and optimal control, we further develop an effective optimization algorithm for mitigating network failures.

In the context of epidemic spreading, many intricate dynamical patterns can emerge due to the cooperation of different types of pathogens or the interaction between the disease spread and other failure propagation mechanism. To unravel such patterns, simulation frameworks are usually adopted, but they are computationally demanding on big networks and subject to large statistical uncertainty. Here, we study the two-layer spreading processes on unidirectionally dependent networks, where the spreading infection of diseases or malware in one layer can trigger cascading failures in another layer and lead to secondary disasters, e.g., disrupting public services, supply chains, or power distribution. We utilize a dynamic message-passing method to devise efficient algorithms for inferring the system states, which allows one to investigate systematically the nature of complex intertwined spreading processes and evaluate their impact. Based on such dynamic message-passing framework and optimal control, we further develop an effective optimization algorithm for mitigating network failures.

Goose parvovirus (GPV) is a pathogen that causes Derzsy’s disease, which causes severe economic losses to China’s waterfowl breeding industry. The mortality rate of Derzsy’s disease reaches 90% in the 7 days. Lactobacillus casei (L. casei) is a common candidate vector in oral vaccines. It is widely used to prevent animal infectious diseases, parasites and tumors. L. casei is a promising vector for developing novel live bacterial vaccine vectors. In this study, L. casei was used as a presentation vector to construct a recombinant Lactobacillus casei expressing the VP3 gene of GPV, which was used as an oral vaccine to immunize goslings. and to evaluate its immunogenicity. In this study, we constructed recombinant Lactobacillus casei expressing the VP3 gene of goose parvovirus, and the immunogenicity and protective efficacy of Lc-pPG-612-VP3 were evaluated in SD rats and goslings. After oral immunization, Lc-pPG612-VP3 colonized the intestine for approximately 34 days and had good immunogenicity. The results showed that Lc-pPG-612-VP3 can induce humoral and mucosal immune responses in goslings, and enhance the transcriptional levels of cytokines in various tissues. The challenge experiment showed that the recombinant Lactobacillus casei had a certain preventive effect against Derzsy’s disease (30%), which was the same as the protection efficiency of the commercial vaccine., Parvovirus gusaka (GPV) jest patogen koji uzrokuje Derzsyjevu bolest. Bolest dovodi do velikih ekonomskih šteta u uzgoju vodenih ptica u Kini. Stopa smrtnosti od Derzsyjeve bolesti doseže i do 90% u tjedan dana. Lactobacillus casei (L. casei) zajednički je vektorski kandidat u oralnim cjepivima. Široko se primjenjuje u sprečavanju zaraznih i nametničkih bolesti životinja kao i prevenciji tumora. L. casei je obećavajući vektor u razvoju novih živih bakterijskih vektorskih cjepiva koja se primjerice mogu koristiti za oralnu imunizaciju guščića. U ovom je istraživanju načinjen rekombinantni Lactobacillus casei koji eksprimira gen VP3 parvovirusa gusaka, a imunogenost i zaštitna učinkovitost Lc-pPG-612-VP3 procijenjene su u SD štakora i guščića. Nakon peroralne imunizacije Lc-pPG-612-VP3 kolonizirao je crijevo približno 34 dana i imao je dobru imunogenost. Rezultati su pokazali da Lc-pPG-612-VP3 može inducirati humoralni i mukozni imunosni odgovor u guščića te povisiti transkripcijske razine citokina u različitim tkivima. Na taj je način utvrđeno da je rekombinantni Lactobacillus casei imao stanoviti preventivni učinak protiv Derzsyjeve bolesti (30%), koji je bio jednak učinkovitosti komercijalnog cjepiva u zaštiti od te bolesti.

Speaking out for women's daily needs on social media has become a crucial form of everyday feminism in China. Gender debate naturally intertwines with such feminist advocacy, where users in opposite stances discuss gender-related issues through intense discourse. The complexities of gender debate necessitate a systematic understanding of discursive strategies for achieving effective gender communication that balances civility and constructiveness. To address this problem, we adopted a mixed-methods study to navigate discursive strategies in gender debate, focusing on 38,636 posts and 187,539 comments from two representative cases in China. Through open coding, we identified a comprehensive taxonomy of linguistic strategies in gender debate, capturing five overarching themes including derogation, gender distinction, intensification, mitigation, and cognizance guidance. Further, we applied regression analysis to unveil these strategies' correlations with user participation and response, illustrating the tension between debating tactics and public engagement. We discuss design implications to facilitate feminist advocacy on social media. Content Warning: This paper contains discussions on gender debate that may include swear words and sensitive topics, such as sex, potentially causing discomfort. © 2024 Copyright held by the owner/author(s)

In recent years, large-scale models can be easily scaled to trillions of parameters with sparsely activated mixture-of-experts (MoE), which significantly improves the model quality while only requiring a sub-linear increase in computational costs. However, MoE layers require the input data to be dynamically routed to a particular GPU for computing during distributed training. The highly dynamic property of data routing and high communication costs in MoE make the training system low scaling efficiency on GPU clusters. In this work, we propose an extensible and efficient MoE training system, ScheMoE, which is equipped with several features. 1) ScheMoE provides a generic scheduling framework that allows the communication and computation tasks in training MoE models to be scheduled in an optimal way. 2) ScheMoE integrates our proposed novel all-to-all collective which better utilizes intra- and inter-connect bandwidths. 3) ScheMoE supports easy extensions of customized all-to-all collectives and data compression approaches while enjoying our scheduling algorithm. Extensive experiments are conducted on a 32-GPU cluster and the results show that ScheMoE outperforms existing state-of-the-art MoE systems, Tutel and Faster-MoE, by 9%-30%. © 2024 ACM.

Directed Acyclic Graph (DAG)-based blockchain represents a paradigm shift from conventional blockchains, which has the potential to drastically improve throughput performance through concurrent storage and executions. In practice, however, existing DAG-based blockchains fail to deliver such promises, often with limited throughput, high conflicts, and security vulnerabilities under dynamic workloads. The root causes are their unawareness of the workload characteristics of different workload sizes and skewed access patterns. In this paper, we propose MorphDAG, the first workload-aware DAG-based blockchain that can significantly enhance throughput without compromising security and achieve elastic scaling under realistic workloads. We derive the theoretically optimal degree of storage concurrency to achieve high throughput while retaining system security as the workload size changes, while enabling fine-grained concurrency adjustment that accommodates a Proof-of-Stake (PoS)-based consensus protocol. We develop a dual-mode transaction processing mechanism that effectively resolves the conflicts brought by skewed access. We implement a prototype of MorphDAG and evaluate under real-world workloads. Extensive evaluations demonstrate that MorphDAG improves end-to-end throughput by up to 2.3× and 2.4× over state-of-the-art DAG-based blockchain systems AdaptChain and OHIE, respectively. Authors

Federated learning allows multiple clients to collaboratively train a global model without revealing their private data. Despite its success in many applications, it remains a challenge to prevent malicious clients to corrupt the global model through uploading incorrect model updates. Hence, one critical issue arises in how to validate the training is truly conducted on legitimate neural networks. To address the issue, we propose VPNNT, a zero-knowledge proof scheme for neural network backpropagation. VPNNT enables each client to prove to others that the model updates (gradients) are indeed calculated on the global model of the previous round, without leaking any information about the client's private training data. Our proof scheme is generally applicable to any type of neural network. Different from conventional verification schemes constructing neural network operations by gate-level circuits, we improve verification efficiency by formulating the training process using custom gates — matrix operations, and apply an optimized linear time zero knowledge protocol for verification. Thanks to the recursive structure of neural network backward propagation, common custom gates are combined in verification thereby reducing prover and verifier costs over conventional zero knowledge proofs. Experimental results show that VPNNT is a lightweighted verification scheme for neural network backpropagation with an improved prove time, verification time and proof size. IEEE

The evolution of Internet-of-Things (IoT) is fostering the use of intelligent controls for energy conservation. Yet, the efficacy of these strategies is largely tied to diverse load forecasting algorithms. Given the significant contribution of heating, ventilation, and air-conditioning (HVAC) systems to global energy consumption, accurate forecasting of HVAC power usage is crucial for improving overall energy efficiency. However, real-world HVAC load forecasting, bolstered by various IoT devices, is complicated by multiple factors: data variability, power load fluctuations, electronic phenomena (e.g., zero drifts), and the increased time complexity and larger model sizes required to manage accumulating historical data. To address these challenges, we first present an in-depth measurement study on the characteristics of HVAC load at a minute scale based on HVAC data collected in six locations. We propose HALO, a transformer-based framework specifically designed for forecasting HVAC load. HALO incorporates an adaptive data pre-processing stage and a local-global-scale transformer-based load forecasting stage, enabling precise forecasting of HVAC load and optimization of energy utilization. Evaluation based on real-world data traces from a prototype application demonstrates that the proposed framework significantly outperforms existing models. IEEE

Social media images, curated or casual, have become a crucial component of communicating situational information and emotions during health crises. Despite its prevalence and significance in informational dissemination and emotional connection, there lacks a comprehensive understanding of visual crisis communication in the aftermath of a pandemic which is characterized by uncertain local situations and emotional fatigue. To fill this gap, this work collected 345, 423 crisis-related posts and 65, 376 original images during the Xi'an COVID-19 local outbreak in China, and adopted a mixed-methods approach to understanding themes, goals, and strategies of crisis imagery. Image clustering captured the diversity of visual themes during the outbreak, such as text images embedding authoritative guidelines and “visual diaries” recording and sharing the quarantine life. Through text classification of the post that visuals were situated in, we found that different visual themes highly correlated with the informational and emotional goals of the post text, such as adopting text images to convey the latest policies and sharing food images to express anxiety. We further unpacked nuanced strategies of crisis image use through inductive coding, such as signifying authority and triggering empathy. We discuss the opportunities and challenges of crisis imagery and provide design implications to facilitate effective visual crisis communication. © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

Flow of typical non-Newtonian fluids such as cement grouts can experience different regimes as the Reynolds number (Re) changes, understanding of which is important for design and operation of rock grouting in various rock engineering applications. Here, flow regimes of representative non-Newtonian fluids, i.e., Bingham and Herschel–Bulkley (H–B) fluids, are numerically investigated with experimental validations. Three tensile rock fracture surfaces originated from a fine-grained sandstone, a medium-grained sandstone and a medium-grained granite samples are used to create rough-walled fracture models with variable aperture structures. Flow of groundwater, Bingham and H–B fluids through these fractures is numerically simulated respectively, by solving the full mass and momentum conservation equations with the Re ranging from 0.01 to 1000. The regimes for these fluids flowing through the fractures are characterized. Laboratory flow tests are conducted in a cylindrical granite fracture sample to verify the characterized flow regimes. The results reveal important differences of flow regimes between Newtonian and non-Newtonian fluids. Specifically, the transmissivity for water flow is constant when Re is relatively small until the Re reaches certain critical values; the transmissivity for Bingham and H–B fluids flow increases with increasing Re until asymptotically reaches certain peak values, followed by a descending stage when Re is relatively large. The critical (water) and peak (Bingham and H–B fluids) values are affected by surface roughness, that is, a rougher surface results in smaller critical and peak values as well as greater discrepancies compared to the analytical solutions based on the smoothed parallel plates model. These results show that a peak or optimum transmissivity is achievable in a specific range of Re (Re = 10–100 for the fractures studied). This new finding can potentially help optimize the injection pressure or flow rate in rock grouting practic, QC 20240719

With continuous urbanization and climate warming, increased air temperature and elevated ozone (O3) concentration often co-occur in many urban areas, but we still lack information about the interactive effects of warming and elevated O3 on urban trees. In the present experiment, the single and combined effects of increased air temperature (IT, ambient air temperature + 2 °C) and elevated O3 (EO, ambient air O3 concentrations + 80 ppb) on carbon (C) fixation and allocation in Quercus mongolica and Pinus tabuliformis, which are widely used as street tree species in urban areas of China, were investigated over two consecutive growing seasons by using 13C isotope techniques. The results showed that IT increased biomass, photosynthetic gas exchange parameters and total 13C content of both tree species. Compared to ambient temperature, IT significantly increased the total 13C content labelled by 56.6 % in Q. mongolica and by 31.2 % in P. tabuliformis in 2021. Elevated O3 induced a decrease in biomass and net photosynthetic rate (Pn) in both tree species. Compared to ambient O3, elevated O3 significantly decreased Pn by 52.6 % in Q. mongolica and by 37.4 % in P. tabuliformis in 2020. The treatment EO decreased 13C allocation to roots but increased 13C content and distribution in leaves in Q. mongolica. These findings demonstrated that EO inhibited the growth and photosynthesis of the two tree species. Our results showed that Q. mongolica was more sensitive to IT and EO than P. tabuliformis, but the former has a self-repair mechanism under increased O3 stress as it is able to invest more carbon to repair leaf damage to a certain extent. Our study also found that the total biomass, relative growth rate, Pn and total 13C content remained higher under the combination of IT and EO compared to EO alone, suggesting that moderate warming may mitigate the negative effects of elevated O3 stress to some extent.

Imaging markers of cerebral small vessel disease provide valuable information on brain health, but their manual assessment is time-consuming and hampered by substantial intra-and interrater variability. Automated rating may benefit biomedical research, as well as clinical assessment, but diagnostic reliability of existing algorithms is unknown. Here, we present the results of the VAscular Lesions DetectiOn and Segmentation (Where is VALDO?) challenge that was run as a satellite event at the international conference on Medical Image Computing and Computer Aided Intervention (MICCAI) 2021. This challenge aimed to promote the development of methods for automated detection and segmentation of small and sparse imaging markers of cerebral small vessel disease, namely enlarged perivascular spaces (EPVS) (Task 1), cerebral microbleeds (Task 2) and lacunes of presumed vascular origin (Task 3) while leveraging weak and noisy labels. Overall, 12 teams participated in the challenge proposing solutions for one or more tasks (4 for Task 1-EPVS, 9 for Task 2-Microbleeds and 6 for Task 3Lacunes). Multi-cohort data was used in both training and evaluation. Results showed a large variability in performance both across teams and across tasks, with promising results notably for Task 1-EPVS and Task 2-Microbleeds and not practically useful results yet for Task 3-Lacunes. It also highlighted the performance inconsistency across cases that may deter use at an individual level, while still proving useful at a population level.

In recent years, the growing pervasiveness of wearable technology has created new opportunities for medical and emergency rescue operations to protect users' health and safety, such as cost-effective medical solutions, more convenient healthcare and quick hospital treatments, which make it easier for the Internet of Medical Things (IoMT) to evolve. The study first presents an overview of the IoMT before introducing the IoMT architecture. Later, it portrays an overview of the core technologies of the IoMT, including cloud computing, big data and artificial intelligence, and it elucidates their utilization within the healthcare system. Further, several emerging challenges, such as cost-effectiveness, security, privacy, accuracy and power consumption, are discussed, and potential solutions for these challenges are also suggested.

Erosion and deposition induced by overtopping flow strongly impacts the breaching evolution of landslide dams. However, these two processes are not yet sufficiently captured in current dam breaching models. In this paper, we simulate the Erosion and Deposition process during overtopping dam breaching using a novel Smooth Particle Hydrodynamics (ED-SPH) method. A new multi-function boundary is proposed to prevent the penetration of water and soil particles and ensure steady inflows. Dam break and movable erosion scenarios are simulated as a means to validate the model. Dam breach simulations using the ED-SPH method reveal that the proposed model can capture the complex dam soil erosion, entrainment, and depositions behaviors. The soil deposition hinders the eroded particle movement and reduces the water velocity at the water-soil interface. If soil deposition is not considered, as in many existing models, the breaching time decreases and the outburst flood discharge is magnified. We further investigate the re-erosion of deposited soils by varying the re-erosion coefficient. With decreasing the re-erosion coefficient, the peak discharge increases whereas the residual dam height decreases due to the blockage by the deposited soils.

Federated learning (FL) has emerged as a prominent distributed learning paradigm, enabling collaborative training of neural network models across local devices with raw data stay local. However, FL systems often encounter significant challenges due to data heterogeneity. Specifically, the non-IID dataset in FL systems substantially slows down the convergence speed during training and adversely impacts the accuracy of the final model. In our paper, we introduce a novel client selection framework that judiciously leverages correlations across local datasets to accelerate training. Our framework first employs a lightweight locality-sensitive hashing algorithm to extract client features while respecting data privacy and incurring minimal overhead. We then design a novel Neural Contextual Combinatorial Bandit (NCCB) algorithm to establish relationships between client features and rewards, enabling intelligent selection of client combinations. We theoretically prove that our proposed NCCB has a bounded regret. Extensive experiments on real-world datasets further demonstrate that our framework surpasses state-of-the-art solutions, resulting in a 50% reduction in training time and a 17% increase in final model accuracy, closing to the performance in the ideal IID case.

This paper proposes a new closed-form design of wideband infinite impulse response digital integrators using numerical integration rules that constructed via trigonometric interpolation. The proposed method can be used to design integrators satisfying prescribed frequency-domain specifications. Design examples with various specifications are given. Simulation comparisons with existing methods are presented to demonstrate the effectiveness of the new design method. The proposed method is experimentally verified on two real applications, Rogowski current transducer and strapdown inertial navigation system. The results show that our method can efficiently improve the integration bandwidth under low sampling rates compared to existing ones, and hence is attractive for resource-constrained applications that require wideband integration.

Continuous authentication is viewed to be increasingly important for mobile devices, which store a wide range of private data and sensitive information of users. Traditional continuous authentication methods need user inputs (e.g. typing, sliding). In this work, we present MAUTH, a zero-effect continuous authentication scheme for mobile devices. With the built-in motion sensors on commercial off-the-shelf (COTS) devices, MAUTH can continuously extract, classify and verify the unique tremor features of users on how their body intrinsically shakes during the normal use of such devices. As a result, it is extremely difficult if not impossible to reproduce the same set of tremors as individuals differ in their muscle development. We implement MAUTH as a software on Android-based smartphones, which demonstrates that MAUTH is light-weight and unobtrusive to its users. We conduct extensive real-world experiments and trace-driven simulations in controlled and uncontrolled environments on 21 volunteers. The results show that MAUTH is difficult to counterfeit and achieves a low average false positive rate (FPR) of 6.73% under real-world spoofing attacks. Moreover, MAUTH is comfortable to use and can achieve a low average false negative rate (FNR) of 2.2% during uncontrolled and continuous usage of devices, leveraging isolation-forest-based classifiers trained with only 40 training samples.

Distributed applications that utilize heterogeneous blockchain systems have the potential to be widely deployed. In such applications, users from different blockchains can transact with one another through cross-chain transactions . There are two essential features of particular relevance for those applications during cross-chain transactions: the atomicity in that either all or none of the blockchains involved confirm a cross-chain transaction, the confidentiality in that a blockchain involved in a cross-chain transaction is only accessible for designated users. Existing cross-chain proposals have largely relied on permissioned blockchains to ensure confidentiality. However, we found that failures could occur when reading or writing information during transaction confirmations across permissioned blockchains, namely read/write (r/w) failures, which can lead to the violation of atomicity. In this paper, we propose a novel mechanism, Unity , to ensure both atomicity and confidentiality of cross-chain transactions under r/w failures by leveraging permissioned blockchains. When failures occur in reading or writing data, Unity classifies the data into two categories based on its status - whether data is the latest version or not, and presents different solutions for atomicity. Specifically, when data is not the latest, we design a four-phase-commit protocol4pc, in which consensus on confirming or aborting a cross-chain transaction can be achieved. If data is the latest when r/w failures occur, we propose a smart contract based solution SSC. We examine the effectiveness of Unity theoretically and through experiments. With a failure probability of 0.7, Unity achieves 98% more atomic cross-chain transactions when compared with the state-of-the-art cross-chain platform, Hyperservice.

Xin Xue,1,2 Li-Bo Zhao,3 Zhe Zhao,3 Wei-Hao Xu,4 Wei-Meng Cai,1 Shao-Hua Chen,1 Tian-Jiao Li,5 Ting-Yu Nie,5 Dong Rui,1 Xiao-Shun Qian,1 Lin Liu1 1Department of Pulmonary and Critical Care Medicine, Second Medical Center of Chinese PLA General Hospital, Beijing, People’s Republic of China; 2National Clinical Research Center for Geriatric Diseases, Chinese PLA General Hospital, Beijing, People’s Republic of China; 3Department of Vasculocardiology, Second Medical Center of Chinese PLA General Hospital, Beijing, People’s Republic of China; 4Department of Geriatrics, Guangdong Provincial Geriatrics Institute, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Southern Medical University, Guangzhou, People’s Republic of China; 5Medical College, Yan’ an University, Yan’ an, People’s Republic of ChinaCorrespondence: Lin Liu; Xiao-Shun Qian, Department of Pulmonary and Critical Care Medicine, Second Medical Center of Chinese PLA General Hospital, Beijing, People’s Republic of China, Tel +86 132 6318 9578 ; +86 135 2127 9287, Email liulin715@qq.com; 13521279287@163.comBackground: The concomitant rise in the prevalence of obstructive sleep apnea (OSA) and frailty among the elderly population has been linked to an increase in mortality rates. Despite continuous positive airway pressure (CPAP) being the gold standard treatment for OSA, its impact on incident frailty remains inadequately explored.Methods: In this cohort study, we analyzed data from 1290 patients diagnosed with OSA, aged 60 years and older. A subset of 71 patients who demonstrated high adherence to CPAP therapy were categorized as the CPAP group. Propensity score matching (PSM) was employed at a 1:4 ratio, matching for variables such as age, gender, body mass index (BMI), and sleep apnea-hypopnea index (AHI), to establish a non-CPAP group for comparison. The FRAIL scale was utilized to evaluate the frailty status of participants. Logistic regression analysis examined

Federated learning allows edge devices to collaboratively train a global model without sharing their local private data. Yet, with limited network bandwidth at the edge, communication often becomes a severe bottleneck. In this paper, we find that it is unnecessary to always synchronize the full model in the entire training process, because many parameters already become mature (i.e., stable) prior to model convergence, and can thus be excluded from later synchronizations. This allows us to reduce the communication overhead without compromising the model accuracy. However, challenges are that the local parameters excluded from global synchronization may diverge on different clients, and meanwhile some parameters may stabilize only temporally. To address these challenges, we propose a novel scheme called Adaptive Parameter Freezing (APF), which fixes (freezes) the non-synchronized stable parameters in intermittent periods. Specifically, the freezing periods are tentatively adjusted in an additively-increase and multiplicatively-decrease manner—depending on whether the previously-frozen parameters remain stable in subsequent iterations. We also extend APF into APF# and APF++, which freeze parameters in a more aggressive manner to achieve larger performance benefit for large complex models. We implemented APF and its variants as Python modules with PyTorch, and extensive experiments show that APF can reduce data transfer amount by over 60%. IEEE

Kink oscillations are frequently observed in coronal loops. This work aims to numerically clarify the influence of loop curvature on horizontally and vertically polarized kink oscillations. Working within the framework of ideal MHD, we conduct 3D simulations of axial fundamental kink oscillations in curved density-enhanced loops embedded in a potential magnetic field. Both horizontal and vertical polarizations are examined, and their oscillation frequencies are compared with WKB expectations. We discriminate two different density specifications. In the first (dubbed"uniform-density"), the density is axially uniform and varies continuously in the transverse direction toward a uniform ambient corona. Some further stratification is implemented in the second specification (dubbed"stratified"), allowing us to address the effect of evanescent barriers. Examining the oscillating profiles of the initially perturbed uniform-density loops, we found that the frequencies for both polarizations deviate from the WKB expectation by $\sim 10\%$. In the stratified loop, however, the frequency of the horizontal polarization deviates to a larger extent ($\sim 25\%$). We illustrate the lateral leakage of kink modes through wave tunnelling in 3D simulations, for the first time. Despite this, in both loops, the damping time-to-period ratios are similar and close to the analytical predictions for straight configurations under the thin-tube-thin-boundary (TTTB) assumption. The WKB expectation for straight configurations can reasonably describe the eigenfrequency of kink oscillations only in loops without an asymmetrical cross-loop density profile perpendicular to the oscillating direction. Lateral leakage via wave tunnelling is found to be less efficient than resonant absorption, meaning that the latter remains a robust damping mechanism for kink motions even when loop curvature is included., Comment: accepted for publication in A&A

Kink oscillations are ubiquitously observed in solar coronal loops, their understanding being crucial in the contexts of coronal seismology and atmospheric heating. We study kink modes supported by a straight coronal loop embeded in an asymmetric environment using three-dimensional magnetohydrodynamic (MHD) simulations. We implement the asymmetric effect by setting different exterior densities below and above the loop interior, and initiate the simulation using a kink-like velocity perturbation perpendicular to the loop plane, mimicking the frequently measured horizontally polarized kink modes. We find that the external velocity fields show fan blade structures propagating in the azimuthal direction as a result of the successive excitation of higher azimuthal Fourier modes. Resonant absorption and phase mixing can still occur despite an asymmetric environment, leading to the development of small scales at loop boundaries. These small scales nonetheless develop asymmetrically at the upper and lower boundaries due to the different gradients of the Alfven speed. These findings enrich our understanding of kink modes in coronal loops embedded within an asymmetric environment, providing insights helpful for future high-resolution observations., Comment: Accepted for publication in A&A

Imaging markers of cerebral small vessel disease provide valuable information on brain health, but their manual assessment is time-consuming and hampered by substantial intra- and interrater variability. Automated rating may benefit biomedical research, as well as clinical assessment, but diagnostic reliability of existing algorithms is unknown. Here, we present the results of the VAscular Lesions DetectiOn and Segmentation (Where is VALDO?) challenge that was run as a satellite event at the international conference on Medical Image Computing and Computer Aided Intervention (MICCAI) 2021. This challenge aimed to promote the development of methods for automated detection and segmentation of small and sparse imaging markers of cerebral small vessel disease, namely enlarged perivascular spaces (EPVS) (Task 1), cerebral microbleeds (Task 2) and lacunes of presumed vascular origin (Task 3) while leveraging weak and noisy labels. Overall, 12 teams participated in the challenge proposing solutions for one or more tasks (4 for Task 1-EPVS, 9 for Task 2-Microbleeds and 6 for Task 3-Lacunes). Multi-cohort data was used in both training and evaluation. Results showed a large variability in performance both across teams and across tasks, with promising results notably for Task 1-EPVS and Task 2-Microbleeds and not practically useful results yet for Task 3-Lacunes. It also highlighted the performance inconsistency across cases that may deter use at an individual level, while still proving useful at a population level.

This study is motivated by observations of coordinated transverse displacements in neighboring solar active region loops, addressing specifically how the behavior of kink motions in straight two-tube equilibria is impacted by tube interactions and tube cross-sectional shapes.We work with linear, ideal, pressureless magnetohydrodynamics. Axially standing kink motions are examined as an initial value problem for transversely structured equilibria involving two identical, field-aligned, density-enhanced tubes with elliptic cross-sections (elliptic tubes). Continuously nonuniform layers are implemented around both tube boundaries. We numerically follow the system response to external velocity drivers, largely focusing on the quasi-mode stage of internal flows to derive the pertinent periods and damping times. The periods and damping times we derive for two-circular-tube setups justify available modal results found with the T-matrix approach. Regardless of cross-sectional shapes, our nonuniform layers feature the development of small-scale shears and energy accumulation around Alf\'ven resonances, indicative of resonant absorption and phase-mixing. As with two-circular-tube systems, our configurational symmetries make it still possible to classify lower-order kink motions by the polarization and symmetric properties of the internal flows; hence such mode labels as $S_x$ and $A_x$. However, the periods and damping times for two-elliptic-tube setups further depend on cross-sectional aspect ratios, with $A_x$ motions occasionally damped less rapidly than $S_x$ motions. We find uncertainties up to $\sim 20\%$ ($\sim 50\%$) for the axial Alfven time (the inhomogeneity lengthscale) if the periods (damping times) computed for two-elliptic-tube setups are seismologically inverted with canonical theories for isolated circular tubes., Comment: Accepted for publication in A&A

Extracellular vesicles (EVs), through their complex cargo, can reflect the state of their cell of origin and change the functions and phenotypes of other cells. These features indicate strong biomarker and therapeutic potential and have generated broad interest, as evidenced by the steady year-on-year increase in the numbers of scientific publications about EVs. Important advances have been made in EV metrology and in understanding and applying EV biology. However, hurdles remain to realising the potential of EVs in domains ranging from basic biology to clinical applications due to challenges in EV nomenclature, separation from non-vesicular extracellular particles, characterisation and functional studies. To address the challenges and opportunities in this rapidly evolving field, the International Society for Extracellular Vesicles (ISEV) updates its ‘Minimal Information for Studies of Extracellular Vesicles’, which was first published in 2014 and then in 2018 as MISEV2014 and MISEV2018, respectively. The goal of the current document, MISEV2023, is to provide researchers with an updated snapshot of available approaches and their advantages and limitations for production, separation and characterisation of EVs from multiple sources, including cell culture, body fluids and solid tissues. In addition to presenting the latest state of the art in basic principles of EV research, this document also covers advanced techniques and approaches that are currently expanding the boundaries of the field. MISEV2023 also includes new sections on EV release and uptake and a brief discussion of in vivo approaches to study EVs. Compiling feedback from ISEV expert task forces and more than 1000 researchers, this document conveys the current state of EV research to facilitate robust scientific discoveries and move the field forward even more rapidly., Extracellular vesicles (EVs), through their complex cargo, can reflect the state of their cell of origin and change the functions and phenotypes of other cells. These features indicate strong biomarker and therapeutic potential and have generated broad interest, as evidenced by the steady year-on-year increase in the numbers of scientific publications about EVs. Important advances have been made in EV metrology and in understanding and applying EV biology. However, hurdles remain to realising the potential of EVs in domains ranging from basic biology to clinical applications due to challenges in EV nomenclature, separation from non-vesicular extracellular particles, characterisation and functional studies. To address the challenges and opportunities in this rapidly evolving field, the International Society for Extracellular Vesicles (ISEV) updates its ‘Minimal Information for Studies of Extracellular Vesicles’, which was first published in 2014 and then in 2018 as MISEV2014 and MISEV2018, respectively. The goal of the current document, MISEV2023, is to provide researchers with an updated snapshot of available approaches and their advantages and limitations for production, separation and characterisation of EVs from multiple sources, including cell culture, body fluids and solid tissues. In addition to presenting the latest state of the art in basic principles of EV research, this document also covers advanced techniques and approaches that are currently expanding the boundaries of the field. MISEV2023 also includes new sections on EV release and uptake and a brief discussion of in vivo approaches to study EVs. Compiling feedback from ISEV expert task forces and more than 1000 researchers, this document conveys the current state of EV research to facilitate robust scientific discoveries and move the field forward even more rapidly.

Extrachromosomal circular DNA (eccDNA) has recently gained increasing attention due to its significant role in cancer and other pathophysiologic states. The majority of circular DNAs detected by Circle-seq are small-size eccDNAs with enigmatic functions. One major technical hurdle is to synthesize eccDNA for functional identification. Here, we describe CAES (Circle-seq based Artificial EccDNA Synthesis), a promising and reliable method for artificial eccDNA synthesis. Eight eccDNAs carrying different microRNA genes (eccMIR) found in gastric cancer tissues, ranging from 329 bp to 2189 bp in size, were created utilizing the CAES method. Exonuclease V and single restriction-endonuclease digestion identified the circular structure of synthetic eccDNAs. The DNA circularization efficiency afforded by CAES ranged from 15.6% to 31.1%, which was negatively correlated with the eccDNA length. In addition, we demonstrated that CAES-synthesized eccMIRs can express both miRNA-3p and − 5p molecules efficiently independent of a canonical promoter in human cell lines. Further assays proved that these eccMIRs were functional as they were able to repress the luciferase gene containing a miRNA-target sequence in the 3′UTR as well as the endogenous mRNA targets. Finally, kinetics study revealed that eccDNA exhibited a decay rate similar to the standard plasmids and linear DNA in cultured cells. Together, this study offers a rapid and convenient method for Circle-seq users to synthesize artificial eccDNAs. It also demonstrates the promising potential of eccMIR as a bacterial DNA-free vector for safe and robust miRNA overexpression in both basic research and therapeutic applications., Extrachromosomal circular DNA (eccDNA) has recently gained increasing attention due to its significant role in cancer and other pathophysiologic states. The majority of circular DNAs detected by Circle-seq are small-size eccDNAs with enigmatic functions. One major technical hurdle is to synthesize eccDNA for functional identification. Here, we describe CAES (Circle-seq based Artificial EccDNA Synthesis), a promising and reliable method for artificial eccDNA synthesis. Eight eccDNAs carrying different microRNA genes (eccMIR) found in gastric cancer tissues, ranging from 329 bp to 2189 bp in size, were created utilizing the CAES method. Exonuclease V and single restriction-endonuclease digestion identified the circular structure of synthetic eccDNAs. The DNA circularization efficiency afforded by CAES ranged from 15.6% to 31.1%, which was negatively correlated with the eccDNA length. In addition, we demonstrated that CAES-synthesized eccMIRs can express both miRNA-3p and − 5p molecules efficiently independent of a canonical promoter in human cell lines. Further assays proved that these eccMIRs were functional as they were able to repress the luciferase gene containing a miRNA-target sequence in the 3′UTR as well as the endogenous mRNA targets. Finally, kinetics study revealed that eccDNA exhibited a decay rate similar to the standard plasmids and linear DNA in cultured cells. Together, this study offers a rapid and convenient method for Circle-seq users to synthesize artificial eccDNAs. It also demonstrates the promising potential of eccMIR as a bacterial DNA-free vector for safe and robust miRNA overexpression in both basic research and therapeutic applications.

Lihua Zhu,1,* Lingling Zhou,2,3,* Liang Wang,4,* Cunte Chen,2 Jie Qiao,1 Xinran Huang,1 Xiaoyan Su,1 Shaohua Chen,2 Bo Li,2 Xiuli Wu,2 Yangqiu Li2 1Department of Rheumatology and Immunology, First Affiliated Hospital, Jinan University, Guangzhou, 510632, People’s Republic of China; 2Institute of Hematology, School of Medicine, Key Laboratory for Regenerative Medicine of Ministry of Education, Jinan University, Guangzhou, 510632, People’s Republic of China; 3Department of Hematology, Nanfang Hospital, Southern Medical University, Guangzhou, 510632, People’s Republic of China; 4Department of Oncology, First Affiliated Hospital, Jinan University, Guangzhou, 510632, People’s Republic of China*These authors contributed equally to this workCorrespondence: Yangqiu Li, Institute of Hematology, School of Medicine, Key Laboratory for Regenerative Medicine of Ministry of Education, Jinan University, Guangzhou, 510632, People’s Republic of China, Tel +86 20 85226877, Fax +86 20 85226877, Email yangqiuli@hotmail.com Lihua Zhu, Department of Rheumatology and Immunology, First Affiliated Hospital, Jinan University, Guangzhou, 510632, People’s Republic of China, Email zhulihua20121113@163.comBackground: There have been several studies regarding the susceptibility of A20 gene SNPs (rs2230926 and rs5029937) in rheumatoid arthritis (RA). However, little is known about the association between polymorphisms in the A20 promoter and RA. The aim of this study was to investigate the characteristics of A20 promoter polymorphisms and the association between these polymorphisms and clinical significance in Chinese RA patients.Methods: PCR and sequencing were used to identify A20 gene polymorphisms in peripheral blood mononuclear cells (PBMCs) from 123 RA cases and 31 healthy individuals.Results: Only one SNP (rs5029924) in the A20 gene promoter was identified in RA patients and healthy individuals. 6 patients who carried heterozygous rs5029924 (3918C

Extracellular vesicles (EVs), through their complex cargo, can reflect the state of their cell of origin and change the functions and phenotypes of other cells. These features indicate strong biomarker and therapeutic potential and have generated broad interest, as evidenced by the steady year-on-year increase in the numbers of scientific publications about EVs. Important advances have been made in EV metrology and in understanding and applying EV biology. However, hurdles remain to realising the potential of EVs in domains ranging from basic biology to clinical applications due to challenges in EV nomenclature, separation from non-vesicular extracellular particles, characterisation and functional studies. To address the challenges and opportunities in this rapidly evolving field, the International Society for Extracellular Vesicles (ISEV) updates its ‘Minimal Information for Studies of Extracellular Vesicles’, which was first published in 2014 and then in 2018 as MISEV2014 and MISEV2018, respectively. The goal of the current document, MISEV2023, is to provide researchers with an updated snapshot of available approaches and their advantages and limitations for production, separation and characterisation of EVs from multiple sources, including cell culture, body fluids and solid tissues. In addition to presenting the latest state of the art in basic principles of EV research, this document also covers advanced techniques and approaches that are currently expanding the boundaries of the field. MISEV2023 also includes new sections on EV release and uptake and a brief discussion of in vivo approaches to study EVs. Compiling feedback from ISEV expert task forces and more than 1000 researchers, this document conveys the current state of EV research to facilitate robust scientific discoveries and move the field forward even more rapidly., Extracellular vesicles (EVs), through their complex cargo, can reflect the state of their cell of origin and change the functions and phenotypes of other cells. These features indicate strong biomarker and therapeutic potential and have generated broad interest, as evidenced by the steady year-on-year increase in the numbers of scientific publications about EVs. Important advances have been made in EV metrology and in understanding and applying EV biology. However, hurdles remain to realising the potential of EVs in domains ranging from basic biology to clinical applications due to challenges in EV nomenclature, separation from non-vesicular extracellular particles, characterisation and functional studies. To address the challenges and opportunities in this rapidly evolving field, the International Society for Extracellular Vesicles (ISEV) updates its ‘Minimal Information for Studies of Extracellular Vesicles’, which was first published in 2014 and then in 2018 as MISEV2014 and MISEV2018, respectively. The goal of the current document, MISEV2023, is to provide researchers with an updated snapshot of available approaches and their advantages and limitations for production, separation and characterisation of EVs from multiple sources, including cell culture, body fluids and solid tissues. In addition to presenting the latest state of the art in basic principles of EV research, this document also covers advanced techniques and approaches that are currently expanding the boundaries of the field. MISEV2023 also includes new sections on EV release and uptake and a brief discussion of in vivo approaches to study EVs. Compiling feedback from ISEV expert task forces and more than 1000 researchers, this document conveys the current state of EV research to facilitate robust scientific discoveries and move the field forward even more rapidly.

Extrachromosomal circular DNA (eccDNA) has recently gained increasing attention due to its significant role in cancer and other pathophysiologic states. The majority of circular DNAs detected by Circle-seq are small-size eccDNAs with enigmatic functions. One major technical hurdle is to synthesize eccDNA for functional identification. Here, we describe CAES (Circle-seq based Artificial EccDNA Synthesis), a promising and reliable method for artificial eccDNA synthesis. Eight eccDNAs carrying different microRNA genes (eccMIR) found in gastric cancer tissues, ranging from 329 bp to 2189 bp in size, were created utilizing the CAES method. Exonuclease V and single restriction-endonuclease digestion identified the circular structure of synthetic eccDNAs. The DNA circularization efficiency afforded by CAES ranged from 15.6% to 31.1%, which was negatively correlated with the eccDNA length. In addition, we demonstrated that CAES-synthesized eccMIRs can express both miRNA-3p and − 5p molecules efficiently independent of a canonical promoter in human cell lines. Further assays proved that these eccMIRs were functional as they were able to repress the luciferase gene containing a miRNA-target sequence in the 3′UTR as well as the endogenous mRNA targets. Finally, kinetics study revealed that eccDNA exhibited a decay rate similar to the standard plasmids and linear DNA in cultured cells. Together, this study offers a rapid and convenient method for Circle-seq users to synthesize artificial eccDNAs. It also demonstrates the promising potential of eccMIR as a bacterial DNA-free vector for safe and robust miRNA overexpression in both basic research and therapeutic applications., Extrachromosomal circular DNA (eccDNA) has recently gained increasing attention due to its significant role in cancer and other pathophysiologic states. The majority of circular DNAs detected by Circle-seq are small-size eccDNAs with enigmatic functions. One major technical hurdle is to synthesize eccDNA for functional identification. Here, we describe CAES (Circle-seq based Artificial EccDNA Synthesis), a promising and reliable method for artificial eccDNA synthesis. Eight eccDNAs carrying different microRNA genes (eccMIR) found in gastric cancer tissues, ranging from 329 bp to 2189 bp in size, were created utilizing the CAES method. Exonuclease V and single restriction-endonuclease digestion identified the circular structure of synthetic eccDNAs. The DNA circularization efficiency afforded by CAES ranged from 15.6% to 31.1%, which was negatively correlated with the eccDNA length. In addition, we demonstrated that CAES-synthesized eccMIRs can express both miRNA-3p and − 5p molecules efficiently independent of a canonical promoter in human cell lines. Further assays proved that these eccMIRs were functional as they were able to repress the luciferase gene containing a miRNA-target sequence in the 3′UTR as well as the endogenous mRNA targets. Finally, kinetics study revealed that eccDNA exhibited a decay rate similar to the standard plasmids and linear DNA in cultured cells. Together, this study offers a rapid and convenient method for Circle-seq users to synthesize artificial eccDNAs. It also demonstrates the promising potential of eccMIR as a bacterial DNA-free vector for safe and robust miRNA overexpression in both basic research and therapeutic applications.

Federated unlearning has emerged very recently as an attempt to realize "the right to be forgotten" in the context of federated learning. While the current literature is making efforts on designing efficient retraining or approximate unlearning approaches, they ignore the information leakage risks brought by the discrepancy between the models before and after unlearning. In this paper, we perform a comprehensive review of prior studies on federated unlearning and privacy leakage from model updating. We propose new taxonomies to categorize and summarize the state-of-the-art federated unlearning algorithms. We present our findings on the inherent vulnerability to inference attacks of the federated unlearning paradigm and summarize defense techniques with the potential of preventing information leakage. Finally, we suggest a privacy preserving federated unlearning framework with promising research directions to facilitate further studies as future work.

Previous multi-task dense prediction methods based on the Mixture of Experts (MoE) have received great performance but they neglect the importance of explicitly modeling the global relations among all tasks. In this paper, we present a novel decoder-focused method for multi-task dense prediction, called Mixture-of-Low-Rank-Experts (MLoRE). To model the global task relationships, MLoRE adds a generic convolution path to the original MoE structure, where each task feature can go through this path for explicit parameter sharing. Furthermore, to control the parameters and computational cost brought by the increase in the number of experts, we take inspiration from LoRA and propose to leverage the low-rank format of a vanilla convolution in the expert network. Since the low-rank experts have fewer parameters and can be dynamically parameterized into the generic convolution, the parameters and computational cost do not change much with the increase of experts. Benefiting from this design, we increase the number of experts and its reception field to enlarge the representation capacity, facilitating multiple dense tasks learning in a unified network. Extensive experiments on the PASCAL-Context and NYUD-v2 benchmarks show that our MLoRE achieves superior performance compared to previous state-of-the-art methods on all metrics. Our code is available at https://github.com/YuqiYang213/MLoRE., Comment: Accepted at CVPR 2024

The rapid proliferation of large language models (LLMs) such as GPT-4 and Gemini underscores the intense demand for resources during their training processes, posing significant challenges due to substantial computational and environmental costs. To alleviate this issue, we propose checkpoint merging in pretraining LLM. This method utilizes LLM checkpoints with shared training trajectories, and is rooted in an extensive search space exploration for the best merging weight via Bayesian optimization. Through various experiments, we demonstrate that: (1) Our proposed methodology exhibits the capacity to augment pretraining, presenting an opportunity akin to obtaining substantial benefits at minimal cost; (2) Our proposed methodology, despite requiring a given held-out dataset, still demonstrates robust generalization capabilities across diverse domains, a pivotal aspect in pretraining.

Speaking out for women's daily needs on social media has become a crucial form of everyday feminism in China. Gender debate naturally intertwines with such feminist advocacy, where users in opposite stances discuss gender-related issues through intense discourse. The complexities of gender debate necessitate a systematic understanding of discursive strategies for achieving effective gender communication that balances civility and constructiveness. To address this problem, we adopted a mixed-methods study to navigate discursive strategies in gender debate, focusing on 38,636 posts and 187,539 comments from two representative cases in China. Through open coding, we identified a comprehensive taxonomy of linguistic strategies in gender debate, capturing five overarching themes including derogation, gender distinction, intensification, mitigation, and cognizance guidance. Further, we applied regression analysis to unveil these strategies' correlations with user participation and response, illustrating the tension between debating tactics and public engagement. We discuss design implications to facilitate feminist advocacy on social media., Comment: 19 pages, 3 figures, In Proceedings of the CHI Conference on Human Factors in Computing Systems (CHI '24)

Compressing high-capability Large Language Models (LLMs) has emerged as a favored strategy for resource-efficient inferences. While state-of-the-art (SoTA) compression methods boast impressive advancements in preserving benign task performance, the potential risks of compression in terms of safety and trustworthiness have been largely neglected. This study conducts the first, thorough evaluation of three (3) leading LLMs using five (5) SoTA compression techniques across eight (8) trustworthiness dimensions. Our experiments highlight the intricate interplay between compression and trustworthiness, revealing some interesting patterns. We find that quantization is currently a more effective approach than pruning in achieving efficiency and trustworthiness simultaneously. For instance, a 4-bit quantized model retains the trustworthiness of its original counterpart, but model pruning significantly degrades trustworthiness, even at 50% sparsity. Moreover, employing quantization within a moderate bit range could unexpectedly improve certain trustworthiness dimensions such as ethics and fairness. Conversely, extreme quantization to very low bit levels (3 bits) tends to reduce trustworthiness significantly. This increased risk cannot be uncovered by looking at benign performance alone, in turn, mandating comprehensive trustworthiness evaluation in practice. These findings culminate in practical recommendations for simultaneously achieving high utility, efficiency, and trustworthiness in LLMs. Code and models are available at https://decoding-comp-trust.github.io., Comment: Accepted to ICML'24

We introduce the Proportional Payoff Allocation Game (PPA-Game) to model how agents, akin to content creators on platforms like YouTube and TikTok, compete for divisible resources and consumers' attention. Payoffs are allocated to agents based on heterogeneous weights, reflecting the diversity in content quality among creators. Our analysis reveals that although a pure Nash equilibrium (PNE) is not guaranteed in every scenario, it is commonly observed, with its absence being rare in our simulations. Beyond analyzing static payoffs, we further discuss the agents' online learning about resource payoffs by integrating a multi-player multi-armed bandit framework. We propose an online algorithm facilitating each agent's maximization of cumulative payoffs over $T$ rounds. Theoretically, we establish that the regret of any agent is bounded by $O(\log^{1 + \eta} T)$ for any $\eta > 0$. Empirical results further validate the effectiveness of our approach.

Current federated learning (FL) approaches view decentralized training data as a single table, divided among participants either horizontally (by rows) or vertically (by columns). However, these approaches are inadequate for handling distributed relational tables across databases. This scenario requires intricate SQL operations like joins and unions to obtain the training data, which is either costly or restricted by privacy concerns. This raises the question: can we directly run FL on distributed relational tables? In this paper, we formalize this problem as relational federated learning (RFL). We propose TablePuppet, a generic framework for RFL that decomposes the learning process into two steps: (1) learning over join (LoJ) followed by (2) learning over union (LoU). In a nutshell, LoJ pushes learning down onto the vertical tables being joined, and LoU further pushes learning down onto the horizontal partitions of each vertical table. TablePuppet incorporates computation/communication optimizations to deal with the duplicate tuples introduced by joins, as well as differential privacy (DP) to protect against both feature and label leakages. We demonstrate the efficiency of TablePuppet in combination with two widely-used ML training algorithms, stochastic gradient descent (SGD) and alternating direction method of multipliers (ADMM), and compare their computation/communication complexity. We evaluate the SGD/ADMM algorithms developed atop TablePuppet by training diverse ML models. Our experimental results show that TablePuppet achieves model accuracy comparable to the centralized baselines running directly atop the SQL results. Moreover, ADMM takes less communication time than SGD to converge to similar model accuracy., Comment: 14 pages, 8 figures

Estimating the individuals' potential response to varying treatment doses is crucial for decision-making in areas such as precision medicine and management science. Most recent studies predict counterfactual outcomes by learning a covariate representation that is independent of the treatment variable. However, such independence constraints neglect much of the covariate information that is useful for counterfactual prediction, especially when the treatment variables are continuous. To tackle the above issue, in this paper, we first theoretically demonstrate the importance of the balancing and prognostic representations for unbiased estimation of the heterogeneous dose-response curves, that is, the learned representations are constrained to satisfy the conditional independence between the covariates and both of the treatment variables and the potential responses. Based on this, we propose a novel Contrastive balancing Representation learning Network using a partial distance measure, called CRNet, for estimating the heterogeneous dose-response curves without losing the continuity of treatments. Extensive experiments are conducted on synthetic and real-world datasets demonstrating that our proposal significantly outperforms previous methods.