Herein, a series of viscosity sensitive Orange and deep red phosphorescent cyclometalated platinum (II) based on coordination bond rotation have been designed and utilized as phosphorescence lifetime imaging (PLIM) probe for nucleolar and mitochondrial viscosity for the first time. All the complexes with imidazole ligands display desirable linear relationships between microsecond-scale lifetimes (0.63 to 27.93 μs) and viscosity over a broad range (0.55 to 950 cP). Among them, green and orange red emissive Pt2 and Pt3 could rapidly penetrate and accumulate in nucleolus within 30 min. In addition, Pt4 shows a significantly enhanced deep-red emission intensity with 53.20-fold increase and greater variation in lifetimes (1.07 to 27.93 μs) with an introduction of a rotating DPA moiety into ligand of 2-Phenylquinoline ring, which may endow Pt4 with the ability of targeting and locating in mitochondria specifically. Moreover, PLIM shows the viscosity of Nucleoli increases significantly during cisplatin-induced apoptosis, and a significant enhanced viscosity of mitochondria were also observed in ferroptosis in HeLa cell. Our work provides a new strategy of constructing viscosity sensitive PLIM probes, which can target different subcellular organelle and track the alternations in viscosity quantitatively during process of cellular dysfunction and ferroptosis in living cancer cell in real time.

The rapid development of multimodal transportation poses dual challenges for secure cross-modal trajectory data management. Addressing the core issues of spatial structure distortion and spatiotemporal fragmentation in existing methods, which result in reduced usability and credibility of the generated trajectories. This paper proposes trajectory rule-aware privacy guardian (TRAP-Guard), a dual-layer road network-constrained privacy preservation model. The model utilizes sequential path variational autoencoder based on variational autoencoder architecture with path completion to effectively improve the spatial topological consistency of the generated trajectories. During the trajectory point-level perturbation phase, by integrating the global candidate set of the road network, the exponential mechanism, and temporal constraints, fine-grained completion satisfying graph-geo-indistinguishability is achieved. Experimental validation on real-world public transportation datasets confirms that TRAP-Guard significantly enhances spatial rationality, temporal consistency, and overall utility of privacy-preserved trajectories.

The design of bioactive functional molecules that respond to endogenous stimuli and clarify the underlying reaction mechanisms is extremely important. Herein, a series of highly turn-on responsive phosphorescent cyclometalated platinum(II) complexes with planar ligands containing different electron-withdrawing or donating substituents were rationally designed and synthesized. The substituents on the planar ligand can significantly influence the interaction of complexes on anion turn-on emissive recognition and binding with proteins. As Pt3a with C<^>N ligand of methyl substituent shows the highest enhancement in turn on sensing towards ClO4 -, I-, and Br- , among 14 kinds of anions, moreover, Pt complex and anion can self-assemble dynamically into red emissive nanomaterials with different morphologies, such as nanonets and nanowires, which were investigated by 1H NMR spectra variation, DLS assay, TEM, SEM and Density functional theory (DFT). Pt3b shows the strongest interaction with BSA, it can distinguish BSA and HSA, and could be used to detect proteinuria among samples by a turn-on green emission. Moreover, Pt3c barely suffers interference from anion or protein, shows a highest cytotoxicity, and can effectively induce necroptosis of HeLa by up-regulating the expression of key proteins. These complexes may show a high potential in developing effective theranostic agents in clinical.

Herein, three pairs of cationic cyclometallated platinum (II) complexes with two distinct types of C^N ligands was rationally designed and synthesized, which contains three kinds of N^N ligands of acetonitrile, metronidazole and histidine. Their cation and anion induced green and orange phosphorescent response, interaction with biomolecules and antitumor activity were studied. It revealed that the cation and anion fluorescent recognition of Pt (II) complexes are highly correlated with N^N than C^N ligand, respectively. Pt complexes of CH3CN and Pt1-his shows best responsive turn on emission on biomolecules of BSA, DNA and RNA. MTT assay demonstrated that Pt1-Me and Pt1-Me2 shows highest cytotoxicity, even higher than that of cisplatin, while rest complexes barely show cytotoxicity. Pt1-Me efficiently entered the cells via an energy-dependent pathway, and it could induce effective apoptosis on Hela. Molecular docking revealed the hydrogen-bonding binding mode of Pt1-CN with BSA, and interaction of Pt1-Me with DNA. Our study has provided a new solution and experimental data for developing novel anticancer theranostic agents.

Platinum-based anticancer drugs represent a cornerstone of clinical cancer chemotherapy, yet their clinical application is severely limited by inherent drug resistance and systemic toxic side effects. Macrocyclic compounds, such as cucurbiturils, cyclodextrins, crown ethers, calixarenes and pillararenes, possess unique cavity structures and excellent molecular recognition capabilities, offering promising strategies for optimizing the chemotherapeutic efficacy of platinum-based drugs. This review summarizes the latest research progress in supramolecular systems constructed from macrocyclic compounds and platinum-based drugs via host-guest interactions. It focuses on the roles of such supramolecular complexes in improving drug solubility, stability, tumor targeting, and controlled release performance, as well as their potential to overcome drug resistance and reduce toxicity. Furthermore, the applications of functionalized macrocyclic compounds in intelligent drug delivery and combination therapy are discussed, and future directions, including novel macrocyclic carriers, multifunctional nanoplatforms, and combination with immunotherapy and gene therapy, are prospected. This review aims to provide useful references for the development of precise and personalized applications of platinum-based drugs in cancer therapy.

For the multimodal sentiment analysis of commentary videos, different modalities contain target information with varying degrees of contribution. Existing methods tend to focus on mining textual modalities that contribute more, which inhibits modalities that contribute less during training, thereby weakening the effective information of modalities that contribute less during fusion. Hence, a modal-correlated generative adversarial network (MGAN) for the multimodal sentiment analysis of commentary videos is proposed, which balances modal contributions by differential complementation. First, temporal-based contextual features are extracted using long short-term memory to accommodate the characteristics of different modalities. The textual modality is identified as the primary contributing modality and the other two modalities as secondary contributing modalities. Second, an MGAN is proposed to generate complementary features with the same distribution as the secondary contributing modalities by correlating between the primary and secondary contributing modalities. To further mine information regarding the discrepancies between the primary and secondary contributing modalities, propose a semantic filtering strategy based on feature similarity. It obtains cross-modality-correlated contextual features by establishing a cross-modal contextual attention mechanism; based on these features, the cross-modal similarity is computed. Additionally, a gating mechanism is established to filter out weakly relevant semantics to enhance the proportion of secondary contributing modalities that differ from the primary contributing modalities and generate more purely differential complementary features that conform to sentiment polarity. Finally, multimodal semantic and differential complementary features are fused to fully utilize the discrepancies in target information contained among the modalities to complement the fusion-model information. Experimental results show that the proposed model performs better on multiple datasets, with an average increase of 3.45 in accuracy compared with the optimal baseline method, particularly on CMU-MOSEI, where the model achieves an accuracy level and F1 score of 92.5 % and 93.8 %, respectively. These results validate the effectiveness of proposed model for multiple datasets.

Herein, a series of orange and deep red phosphorescent cyclometalated Pt(II) complexes with thiazole- or benzothiophene- based ligands were rationally designed. Notably, thiazole-containing complexes exhibit aggregation-caused quenching (ACQ), whereas benzothiophene analogues (Pt-2a-c) display aggregation-induced emission (AIE) with a remarkable red-shift (Δλ = 207-220 nm). Upon addition of biologically relevant anions (Br-, HSO3-, ClO4-) Pt-1a and Pt-2b undergo dynamic self-assembly, which not only modulates the Pt-Pt interactions resulting in distinct phosphorescent emissions but also induces morphological transformation into needle-shaped nanorods, nanotubes, and nano-blocks. The self-assemblewas systematically characterized by 1H NMR, DLS, SEM, CLSM and DFT. Moreover, Pt-1b can distinguish DNA and RNA simultaneously by turn on emission and effectively induce early apoptosis of HeLa. Our work thus reveals the potential of these Pt(II) complexes in develping novel versatile agents for theranostics in vitro.

Due to the varieties of sentiment expressions, multimodal sentiment analysis for social media requires a comprehensive fusion of image and textual information. However, most of the previous studies have only modeled the inter-modal local or global interactions, ignoring inter-modal global and local co-influences, resulting in insufficient fusion of sentiment information. In addition, the introduction of multiple features may generate more sentiment-irrelevant information, thus leading to a weaker sentiment association of the fusion features. To solve the above issues, we propose a global-local feature fusion network model leveraging transformer-encoder and contrastive learning. Firstly, considering inter-modal global and local co-influences, the model extracts global and local features in the image. Secondly, we propose a cross-modal synchronous fusion transformer-encoder and its simplified version to capture inter-modal global and local consistent features, and combine it with soft self-attention to further enhance inter-modal interaction. On this basis, we utilize multiple contrastive learning to enhance the interactions among multiple fusion features and improve the sentiment associations of multimodal fusion features to assist the final sentiment analysis. Extensive experiments on three public multimodal datasets show that our model can adequately capture inter-modal global-local information interactions and effectively improve sentiment associations, thus demonstrating its validity and superiority.

Multimodal sentiment analysis (MSA) provides a more accurate understanding of human emotional states than unimodal. However, the different modalities are limited by semantic expression in expressing emotion, leading to inconsistency in the importance of unimodal influence on the fused modal sentiment polarity, as well as sentiment polarity biases resulting from the interaction between multiple modalities. This can make MSA less accurate. To address this problem, we propose a two-stage adaptive fusion network (TsAFN) in this paper. The first stage is an adaptive fusion network based on the joint of modal features. Feature extraction is based on Bert and LSTM network. An importance metric adaptive benchmark is presented for proposing a feature planning method to jointly represent multimodal features to form fused modal features, which automatically equalizes the importance of unimodal influence on the fused modal sentiment polarity. The second stage is an adaptive fusion network based on modal interaction. A distance metric adaptive benchmark is defined, based on which a representation reconstruction method is proposed to take into account inter-modal interactions. The relationship and sentiment polarity biases of the modalities are adjusted to reconstruct unimodal sentiment polarity and a more accurate representation of the fused modality. Finally, the loss function is defined and the model is trained on three datasets MOSI, MOSEI, and CH-SIMS. The results of comparative experiments show that TsAFN can achieve better accuracy in MSA.

Road segmentation from high-resolution remote sensing imagery is critical for tasks such as autonomous driving, urban planning, and geographic information systems. However, challenges such as intensity nonuniformity, pixel ambiguity, and the visual similarity between roads and natural features make accurate segmentation difficult. In this article, we propose a road segmentation framework built upon the Mamba architecture, integrating a novel frequency feature compensation (FFC) approach to improve segmentation performance. Specifically, we introduce a progressive FFC method, leveraging wavelet decomposition to capture fine-grained details by separating features into high- and low-frequency components. Multistage features extracted from the Mamba backbone are decomposed using this approach and progressively integrated to compensate for the essential details for accurate road segmentation. We also introduce a wavelet loss (WL) to improve the model's ability to capture fine structural variations in the frequency domain. Furthermore, we develop a spatial perception Mamba block (SPMB) to enhance the capture of spatial relationships. By seamlessly integrating global context and local structures with the selective state-space model and FFC, our framework significantly boosts road segmentation accuracy. Extensive experiments on three publicly available road segmentation datasets demonstrate that our method achieves state-of-the-art performance, surpassing existing approaches in segmenting complex roads.

Herein, ruthenium (II) polypyridine complexes with imidazole and bi-imidazole ligands were first studied as specific anion chromogenic sensors in aqueous media, both in the dark and under light irradiation. Under light, Ru1-2 of ethyl-imidazole ligands showed a better performance in sensing CO32-, HSO3- , and ClO- via an obvious color change that can be detected by the naked eye over other 13 kinds of anions, compared with their performance in the dark. Ru3 and Ru5 of the bidentate ligand could sense four anions of CO32-, HSO3- , ClO-, and HCO3in aqueous media by colorimetry without irradiation. Besides, the specific anions can significantly quench the deep red emission of each complex, and Ru3 and Ru5 can act as emissive "turn-off" anion recognition probe. The paper strips loaded with Ru5 can be used to detect different concentrations of CO32- (2.5-10 mM) in aqueous media under sunlight and ultraviolet light. Moreover, Ru2 as a representative complex in the series shows low cytotoxicity and high cellular uptake under dark condition, and an enhanced phototoxicity within 72 h, which shows good potential to be a PDT and PACT therapy agent.

With the rapid advancement of autonomous driving technology, precise user trajectory data has become critical for vehicles. Current privacy-preserving methodologies predominantly focus on spatial feature extraction while insufficiently addressing temporal dimension vulnerabilities, a security-critical oversight that disrupts the privacy-utility trade-off. We propose a Dual-Layer Temporal Privacy Protection Model (DLTPPM) featuring: (1) A modified Trajectory Chain Variational Autoencoder (TCVAE) combining Long Short-Term Memory (LSTM) with enhanced variational inference for expressive spatiotemporal pattern learning; (2) A Random Forest-Laplace Time Protector (RF-LTP) applying context-aware differential privacy via adaptive temporal noise injection. Experimental validation confirms DLTPPM’s capability to simultaneously enhance privacy protection and data utility, demonstrating significant balanced improvement over conventional approaches.

Over the past few years, multiview attributed graph clustering has achieved promising performance via various data augmentation strategies. However, we observe that the aggregation of node information in multilayer graph autoencoder (GAE) is prone to deviation, especially when edges or node attributes are randomly perturbed. To this end, we innovatively propose a tensor-representation-based multiview attributed graph clustering framework with smooth structure (MV_AGC) to avoid the bias caused by random view construction. Specifically, we first design a novel tensor-product-based high-order graph attention network (GAT) with structural constraints to realize efficient attribute fusion and semantic consistency encoding. By imposing attribute augmentation mechanisms and smooth constraints (SCs) on the proposed high-order graph attention autoencoder simultaneously, MV_AGC effectively eliminates the instability of reconstructed graph structures and learns a more compact node representation during training. In addition, we also theoretically analyze the stronger generality and expressiveness of the proposed tensor-product-based attention mechanism over the classical GAT and establish an intuitive connection between them. Furthermore, to address the performance degradation caused by clustering distribution updating, we further develop a simple yet effective clustering objective function-guided self-optimizing module for the final clustering performance improvement. Experimental results on the six benchmark datasets have demonstrated that our proposed method can achieve state-of-the-art clustering performance.

This study focuses on the synthetic modification of satraplatin through axial position substitution by introducing the bioactive molecule cannabidiol (CBD) at the axial position, leading to novel multifunctional antitumor complexes S1-S5. The results demonstrate that the target complexes exhibit significantly optimized physicochemical properties with improved lipo-hydro partition coefficients, combining both lipophilic and hydrophilic characteristics that favor drug metabolism. Remarkably enhanced in vitro antitumor activity was observed: S3 showed 4.3-fold higher cytotoxicity than satraplatin in HCT-116 cells, with superior efficacy against drug-resistant cells. S5 demonstrated 22-fold greater antiproliferative activity than both parental satraplatin and CBD alone in A549 cells, likely attributable to its haloacetic axial group enhancing mitochondrial targeting for trifunctional synergistic antitumor effects. Mechanistic studies revealed that the complexes are reduced by ascorbic acid in vivo to release both CBD and Pt (II). CBD disrupts redox homeostasis through ROS accumulation and impairs mitochondrial membrane potential, synergizing with Pt (II)-induced DNA damage to promote cancer cell apoptosis. The introduction of bioactive moieties like CBD with excellent biological functions into the axial position of Pt (IV) complexes not only optimizes physicochemical properties but also enriches antitumor functionality. Compared with conventional platinum drugs, these complexes demonstrate significantly enhanced cytotoxicity (S3 is 9.5-fold more toxic than cisplatin in HCT-116 cells; S5 shows 33-fold higher antiproliferative activity in A549 cells) while maintaining superior activity against resistant cells through CBD-Pt (II) synergy. This study provides valuable insights for overcoming limitations of traditional platinum drugs and developing synergistic antitumor agents with enhanced efficacy.

Controlling the size and morphology of supramolecular assemblies is still an enormous challenge for the development of novel functional materials. Herein, cyclometalated platinum(II) complexes were designed and used as effective theragnostic supramolecular nano agents. Benzoquinoline was used as the main (CN)-N-boolean AND ligand, and the increased number of aromatic rings in the (NN)-N-boolean AND ligand of Pt complexes could greatly improve lipophilicity and cytotoxicity towards cancer cells over normal cells. Moreover, the morphology of nanoparticles formed by self-assembly of Pta-d could change from one to three-dimensional (1-3D), which can form nanowires and nano-spheres. Besides, complexes of Pta-c, the number of aromatic rings in the (NN)-N-boolean AND ligand of which does not exceed 4, all exhibit significant aggregation-induced phosphorescent emission (AIPE) with 100- 200-fold enhancement in red phosphoresce intensity recorded in the solvent of Pta-1 and Ptb. Pt-Pt interaction induced by coordination and electrostatic interaction between complex and anions, and a new deep red emission improvement was observed in aqueous solution of Pta and Ptb with the presence of ClO4-, and two similar deep red emissions induced by two different interactions can be told by new emerging MMLCT absorption band. Ptd of dppz ligand exhibits the highest efficiency in inducing apoptosis of HeLa and its anticancer mechanism was studied. Our work aims to promote the fundamental comprehension of the self-assembly behavior of cyclometalated platinum complex with AIPE in vitro and living cells.

Recently, kernelized multi-view clustering has garnered significant attention due to its powerful ability to effectively cluster multi-view data characterized by non-linear structures. However, most existing methods focus more on kernel learning while neglecting graph structure learning in the kernel space, resulting in the creation of an affinity graph that is suboptimal for clustering purposes. To address this issue, this paper proposes a novel kernelized multi-view graph clustering method via graph structure preserving and consensus affinity graph learning. The proposed method first designs a predefined kernel matrix for each individual view. Subsequently, it learns a view-specific candidate affinity graph that preserves both the local and global structures of the input data within the kernel space. Next, a reliable and robust consensus affinity graph, which accurately captures the underlying cluster structure and is resistant to noise, is jointly learned from all the candidate affinity graphs and their shared latent representation. Finally, we integrate graph structure learning in kernel space, shared latent representation learning, and consensus affinity graph learning into a unified framework, enabling them to mutually reinforce each other during iterative optimization. Experiments conducted on benchmark datasets have demonstrated that the proposed method outperforms some state-of-the-art multi-view clustering methods.

This paper presents a dynamic resource orchestration framework for edge computing environments, utilizing multi-agent reinforcement learning (MARL) to enhance resource allocation and task scheduling. The proposed system consists of edge nodes (E), a centralized resource manager (CRM), and communication infrastructure (CI). Edge nodes execute computational tasks at the network’s periphery, while the CRM oversees resource distribution and task assignment using a global system perspective. The CI supports efficient communication among these components. The MARL framework enables collaborative learning among agents, where each agent selects optimal actions—such as resource allocation, task scheduling, and migration—based on system states that include resource availability, task queue lengths, network conditions, and task priorities. A deep Q-network (DQN)-based training approach is employed, allowing agents to maximize cumulative rewards by balancing task completion efficiency, resource utilization, and latency minimization. The proposed framework is evaluated through comprehensive simulations against traditional heuristic-based and static resource allocation methods. Results demonstrate that our MARL-based approach reduces average task completion latency by 12.3

Classical cisplatin-based chemotherapeutic drugs are widely used in clinical practice. In recent years, novel platinum-based antitumor drugs have focused on replacing classical cisplatin-like Pt(II) complexes with relatively inert Pt(IV) prodrugs to overcome drug resistance and reduce toxic side effects. Based on the excellent physiological and pharmacological activities of cannabidiol (CBD), this study designed and synthesized novel Pt(IV) prodrugs W1-W6, which are axial conjugates of cisplatin with CBD and specific active small molecules. These prodrugs demonstrated more significant antitumor activity against tested tumor cell lines. Among them, the multifunctional Pt(IV) prodrug W5, conjugated with CBD and the PDK inhibitor DCA, exhibited excellent activity against both platinum-sensitive and cisplatin-resistant tumor strains. The IC50 value of W5 for the A549R tumor strain was 8.53 ± 0.76 μM, significantly higher than that of the cisplatin group and 3.64 times the activity of CBD alone, demonstrating strong synergistic antitumor activity and potential to overcome cisplatin resistance. W5 is reduced by GSH in A549R cells, releasing CBD and Pt(II). Pt(II) binds to DNA, inducing damage and inhibiting repair, while CBD activates pro-apoptotic proteins, leading to mitochondrial dysfunction. Simultaneously, W5 reduces the levels of ROS scavengers, triggering endoplasmic reticulum dysfunction. These three mechanisms synergistically promote tumor cell apoptosis and overcome drug resistance. This design integrates multiple mechanisms through axial functionalization, breaking through the limitation of traditional platinum drugs targeting DNA alone, and achieves synergistic effects by regulating metabolism and intervening in the immune microenvironment.

In view of the challenges faced by traditional federated learning in edge computing scenarios, such as data heterogeneity and personalized requirements among edge nodes, limiting the adaptability of the global model. So, a comprehensive review of recent advances in personalized federated learning for edge computing scenarios was provided. Firstly, the background and scientific significance of personalized federated learning were elaborated, followed by rigorous analysis of data heterogeneity’s impacts. Secondly, data heterogeneity was formally defined and classified. Subsequently, existing approaches were categorized into five key methodologies: data-based methods, client-side model optimization, server-side aggregation optimization, global architecture optimization, large model, and prototype-based learning methods. Finally, to guide ongoing developments in the field, future trends and outlines potential research directions were explored.

In multimodal sentiment analysis, commentary videos often lack certain sentences or frames, leaving gaps that may contain crucial sentiment cues. Current methods primarily focus on modal fusion, overlooking the uncertainty of missing modalities, which results in underutilized data and less complete and less accurate sentiment analysis. To address these challenges, we propose a prompt-matching synthesis model to handle missing modalities in sentiment analysis. First, we develop unimodal encoders using prompt learning to enhance the model's understanding of inter-modal relationships during feature extraction. Learnable prompts are introduced before textual modalities, while cross-modal prompts are applied to acoustic and visual modalities. Second, we implement bidirectional cross-modal matching to minimize discrepancies among shared features, employing central moment discrepancy loss across multiple modalities. A comparator is designed to infer features based on the absence of one or two modalities, allowing for the synthesis of missing modality features from available data. Finally, the synthesized modal features are integrated with the initial features, optimizing the fusion loss and central moment discrepancy loss to enhance sentiment analysis accuracy. Experimental results demonstrate that our method achieves strong performance on multiple datasets for multimodal sentiment analysis, even with uncertain missing modalities.