Dual prompt personalized federated learning in foundation models.

Personalized federated learning (PFL) has garnered significant attention for its ability to address heterogeneous client data distributions while preserving data privacy. However, when local client data is limited, deep learning models often suffer from insufficient training, leading to suboptimal performance. Foundation models, such as CLIP (Contrastive Language-Image Pretraining), exhibit strong feature extraction capabilities and can alleviate this issue by fine-tuning on limited local data.

Detection of Caco-2 cell absorption rate of Bombyx mori peptide chelated zinc and validation of anti-inflammatory activity based on network pharmacology.

The research of Bombyx mori mainly focuses on the overall brewing of Bombyx mori, there are few studies on Bombyx mori protein and by-products after processing. In order to improving the utilization rate of Bombyx mori, Bombyx mori polypeptide was chelated with inorganic zinc to prepare Bombyx mori polypeptide chelated zinc in this study. Bombyx mori polypeptide was used as the raw material to prepared Bombyx mori polypeptide chelated zinc.

Federated Learning for Human Pose Estimation on Non-IID Data via Gradient Coordination.

Human pose estimation is an important downstream task in computer vision, with significant applications in action recognition and virtual reality. However, data collected in a decentralized manner often exhibit non-independent and identically distributed (non-IID) characteristics, and traditional federated learning aggregation strategies can lead to gradient conflicts that impair model convergence and accuracy. To address this, we propose the Federated Gradient Harmonization aggregation strategy (FedGH), which coordinates update directions by measuring client gradient discrepancies and integrating gradient-projection correction with a parameter-reconstruction mechanism.

Goal-oriented autonomous decision-making for social robots via collaborative interactive inverse reinforcement learning approach.

Since the practical constraints of unknown pedestrian goal information, research on inverse reinforcement learning (IRL) applied to social robots has focused on trajectory planning based on current motion direction, other pedestrians, and obstacles. However, social robots typically have clear navigation goals, and the practicality of the aforementioned methods is debatable. Moreover, trajectory prediction at longer distances also poses significant challenges for such trajectory planning methods.

Study on immunogenicity of recombinant ferritin hemagglutinin of canine distemper virus.

Background: Canine distemper virus (CDV) hemagglutinin protein (H), as one of the surface glycoproteins of the virus, helps attach the virus to the host cell through its interaction with cell receptors, makes hemagglutinin protein a key target for the development of neutralizing antibodies.

Methods: In this study, H protein sequences were analyzed by conservation analysis and prediction of multiple immune cell epitopes, three possible epitope sequences were identified. The epitopes were ligated to Ferritin vector with linkers to construct pET30a-Ferritin-Hemagglutinin-Canine distemper virus (FHCDV).

Modeling the effects of a Shock-and-Kill Treatment for HIV: Latency-Reversing Agents and Natural Killer Cells.

Despite the substantial success of combination antiretroviral therapy (ART) in suppressing HIV replication, achieving a complete cure remains challenging due to the persistence of viral reservoirs. The use of latency-reversing agents (LRAs) combined with natural killer (NK) cells in a "shock-and-kill" strategy has been experimentally confirmed as an effective approach to reducing reservoirs. Here, we utilized an HIV infection mathematical model that incorporates both 'virus-cell' and 'cell-cell' infection modes to assess the dynamic synergy of ART, LRAs, and NK cells.

Self-Healing Superhydrophobic Coatings Based on Fluorinated Polydopamine-Zinc Ion-Graphene Oxide Hybrid: Design and Anticorrosive Applications.

In this work, zinc ions were added to graphene oxide (GO) after it had been modified with polydopamine (PDA). After that, the substance was reacted with 1H,1H,2H,2H-perfluorodecanethiol (PFDT) to produce a two-dimensional (2D) nanomaterial that was hydrophobic. To create self-healing superhydrophobic anticorrosive coatings on AZ31B magnesium alloy, the synthesized nanomaterial was combined with polydimethylsiloxane (PDMS) and applied using a straightforward spray technique.

Quaternized Chitosan-Coated Nano-MOFs as Antigen Nanocarriers with Enhanced Stability and Immunogenicity.

Vaccines are among the greatest achievements of modern medicine, and have saved countless lives. Subunit vaccines are vaccines made from specific parts (subunits) of a pathogen. Compared other types of vaccines, such as inactivated and live-attenuated vaccines, they are generally safer and more suitable for individuals with weakened immune systems.

Food and medicine homology: a potential nutritional intervention strategy for post-acute COVID-19 syndrome.

After recovering from severe acute respiratory syndrome of coronavirus 2 (SARS-CoV-2) acute infection, some patients with corona virus disease 2019 (COVID-19) still are affected by post-acute COVID-19 syndrome (PACS). Traditional Chinese medicine (TCM) has played an important role in the recovery period of COVID-19. As a unique type of Chinese botanical drugs, foods with property of food and medicine homology (FMH) has the dual characteristics of drug and food, which has advantages over usual drugs in safety and daily application.

About Sugar Addiction.

Purpose: Sugar addiction, characterized by excessive cravings for high-sugar foods, poses a significant health challenge in modern society. The parallels between sugar and drug addiction have attracted attention, prompting extensive research on the neurological mechanisms involved.

Method: This review summarizes recent research on sugar addiction and its impact on health.

An integrated all-fiber-based sensing system for monitoring humidity physiological signals.

With the advancement of wearable electronics, flexible sensors have attracted significant attention in health monitoring and personalized medical services. While these sensors exhibit excellent responsiveness for health tracking, they often overlook the importance of skin surface humidity. All-fiber-based sensors exhibit exceptional flexibility, breathability and lightweight nature.

Deciphering thermal degradation pathways of metal halide perovskite thin films under ambient conditions.

Understanding the dynamic lattice collapse pathways under thermal stress is pivotal for designing stable perovskite and improving the power conversion efficiency (PCE) of perovskite solar cells (PSCs). However, the lack of dynamic molecular-level characterization techniques has limited a comprehensive understanding of the degradation pathways involved. In this study, we use temperature-dependent Raman spectroscopy (25-375 °C) coupled with multimodal characterization to investigate A-site cation-dependent degradation pathways in mixed-halide perovskites, specifically FACsPbI (FACs) and FAMAPbI (FAMA).

CosELU: An improved activation deep neural network for EEG signal recognition in stroke patients.

Motor imagery electroencephalograme (MI-EEG) signal recognition algorithms are essential for rehabilitation training of clinical stroke patients. However, EEG signals of stroke patients using traditional activation functions are prone to exploding gradient and vanishing gradient problems in deep learning deep neural network training, leading to a decrease in the recognition rate of EEG signals. In this paper, an adaptive activation function is developed and investigated to address this problem.

Ordered Dynamic Networks Proton Exchange Membrane for Humidity-Resilient High-Performance Fuel Cells.

Maintaining proton conductivity of sulfonated proton exchange membranes (PEMs) under low humidity remains a critical challenge for fuel cell applications. This study presents an innovative interfacial engineering strategy through the integration of perfluorosulfonic acid nanofibers (PFSANF) with hydroxyl-functionalized Tröger's base polymer containing tertiary amine groups (HTB), constructing an ordered dynamic network membrane. The system achieves multi-level performance optimization via synergistic 3D hydrogen-bond networks and acid-base interactions: 1) Nanofiber create high-speed proton transport networks; 2) Hydroxyl and tertiary amine groups cooperatively enhance water retention while creating additional proton-hopping sites.

The sequence kernel association test for the proportional odds model.

Motivation: The Sequence Kernel Association Test (SKAT) and its extensions are the most popular methods for studying the association between phenotypes and a set of single nucleotide polymorphisms. Their practical application is very wide, but most of these methods are designed for continuous and binary phenotypes. Ordered categorical phenotypes are also very common in practice, so there is an urgent need to develop SKAT-type tests for proportional odds model.

Finite-time stability analysis of interacting genetic regulatory networks with spatial diffusion term.

This paper explores the finite-time stable behavior of interacting genetic regulatory networks (GRNs) incorporating spatial diffusion terms under Dirichlet boundary conditions. We propose a coupled model of interacting GRNs and a novel Lyapunov-Krasovskii functional to analyze the dynamic evolution of interactions between the two systems. The lower bound of the time delay can be either zero or non-zero.

MBRSTCformer: a knowledge embedded local-global spatiotemporal transformer for emotion recognition.

Emotion recognition is an essential prerequisite for realizing generalized BCI, which possesses an extensive range of applications in real life. EEG-based emotion recognition has become mainstream due to its real-time mapping of brain emotional activities, so a robust EEG-based emotion recognition model is of great interest. However, most existing deep learning emotion recognition methods treat the EEG signal as a whole feature extraction, which will destroy its local stimulation differences and fail to extract local features of the brain region well.

Nuclear-targeted chiral red-emissive carbon dots for visual detection of leucine aminopeptidase in vitro.

Leucine aminopeptidase (LAP) is a key biomarker for the early diagnosis of drug-induced liver injury (DILI). Traditional detection methods are difficult to meet the actual needs of clinical dynamic monitoring due to problems such as poor selectivity and low sensitivity. To solve this problem, a nuclear targeted dual mode sensor was developed for the sensitive and selective detection of LAP activity on the basis of red emission chiral carbon dots (R-L-CDs).

Monitoring pH in lysosomes of living cells using a pH-sensitive probe based on spirocyclization strategy.

The pH-sensitive fluorescent probe CBTOH, utilizing a spirocyclization mechanism, has been developed for precise pH monitoring and imaging in biological systems. Under acidic conditions (pH 2-3.5), CBTOH undergoes ring-opening, emitting red fluorescence at 611 nm with a 119-fold increase in intensity compared to neutral and alkaline conditions, due to enhanced intramolecular charge transfer upon spirolactam ring cleavage.

Studying the Effect of Receptors Clustering on Hyaluronic Acid Binding with CD44 and the Cell Entry of Hyaluronic Acid Targeting Nanodrugs at Single Molecule/Particle Level.

Receptor-ligand interactions on cell membrane play key roles in many physiological and pathological processes, especially in the nanodrug targeted delivery system. Hyaluronic acid (HA) is the main ligand for binding CD44 to modulate the targeted delivery of nanodrugs; however, the corresponding receptor clustering mechanism remains unclear. The differential effects of HA on CD44 clustering are closely associated with the number of disaccharide units.

Comparative Study on Pervaporation Performance of Polyphosphazene Membranes with Different Fluorine Side Groups for Thiophene/-Heptane Separation.

In recent years, polyphosphazene (POP) membranes have been gaining more and more attention owing to their excellent pervaporation desulfurization performance. To develop new POP membranes, three kinds of POPs with different side groups, Poly[bis(trifluoroethoxy)phosphazene] (PTFEP), Poly[bis(trifluorobutoxy)phosphazene] (PTFBP), and Poly[bis(octafluoropentyloxy)phosphazene] (POFPP), were synthesized. The NMR spectroscopy demonstrated that POPs with a designed structure were successfully prepared.

Low-Field-Driven Domain Wall Motion in Wurtzite Ferroelectrics.

Wurtzite-type nitride ferroelectrics emerge as a breakthrough platform for silicon-compatible nonvolatile memory technology. However, the inherent polarization reversal mechanisms involving diatomic displacements introduce complex domain dynamics and elevate energy barriers, manifesting as excessive coercive fields (E) and pronounced wake-up effects that hinder reliable device operation. Here, these challenges are resolved by enabling the low-field-driven domain wall motion in representative wurtzite ferroelectrics (AlScN).

Revealing the effects of the cell cycle on cellular uptake of nano-drugs at single particle level.

We analyzed the dynamic parameters for nano-drug entry into cells in different stages of the cell cycle, verifying the effects of stiffness and membrane receptor expression level. The cell cycle synchronization inhibitor and cell line were also considered. This report will provide insights into the enhancement of cell entry efficiency of nano-drugs by regulating the cell cycle.

Bayesian Estimation of Partial Functional Tobit Censored Quantile Regression Model.

The information extracted from imaging data has become increasingly important in disease diagnosis as it uncovers associations between imaging features and diseases of interest. This study proposes a partial functional Tobit censored quantile regression (PFTCQR) model to investigate the quantile-specific relationships between the time of incidence of laryngeal cancer and a set of imaging and clinical predictors based on the data collected from a laryngeal cancer study in the Otolaryngology Department of a tertiary hospital in Jilin Province, China. The functional principal component analysis and moment method are employed to estimate the slope and covariance functions of the functional predictors.

Near-Infrared Afterglow Imaging-Guided Surgical Resection and Synergistic Photodynamic-Chemo Therapy of Breast Cancer.

The recurrence rate of cancer following surgical procedures is markedly affected by the degree of tumor removal. During the operation, tumor-targeted imaging is of crucial significance as it assists surgeons in attaining the most comprehensive tumor excision. Herein, a theranostic platform (CDSP NPs) is developed through the assembly of carbon dots (CDs) with paclitaxel prodrugs.