Zhejiang Lab

autophagic responses. Here, we critically discuss current methods of assessing autophagy and the information they can, or cannot, provide. Our ultimate goal is to encourage intellectual and technical innovation in the field.

Y Artificial intelligence (AI) coupled with promising machine learning (ML) techniques well known from computer science is broadly affecting many aspects of various fields including science and technology, industry, and even our day-to-day life. The ML techniques have been developed to analyze high-throughput data with a view to obtaining useful insights, categorizing, predicting, and making evidence-based decisions in novel ways, which will promote the growth of novel applications and fuel the sustainable booming of AI. This paper undertakes a comprehensive survey on the development and application of AI in different aspects of fundamental sciences, including information science, mathematics, medical science, materials science, geoscience, life science, physics, and chemistry. The challenges that each discipline of science meets, and the potentials of AI techniques to handle these challenges, are discussed in detail. Moreover, we shed light on new research trends entailing the integration of AI into each scientific discipline. The aim of this paper is to provide a broad research guideline on fundamental sciences with potential infusion of AI, to help motivate researchers to deeply understand the state-of-the-art applications of AI-based fundamental sciences, and thereby to help promote the continuous development of these fundamental sciences.

Aging is characterized by systemic chronic inflammation, which is accompanied by cellular senescence, immunosenescence, organ dysfunction, and age-related diseases. Given the multidimensional complexity of aging, there is an urgent need for a systematic organization of inflammaging through dimensionality reduction. Factors secreted by senescent cells, known as the senescence-associated secretory phenotype (SASP), promote chronic inflammation and can induce senescence in normal cells. At the same time, chronic inflammation accelerates the senescence of immune cells, resulting in weakened immune function and an inability to clear senescent cells and inflammatory factors, which creates a vicious cycle of inflammation and senescence. Persistently elevated inflammation levels in organs such as the bone marrow, liver, and lungs cannot be eliminated in time, leading to organ damage and aging-related diseases. Therefore, inflammation has been recognized as an endogenous factor in aging, and the elimination of inflammation could be a potential strategy for anti-aging. Here we discuss inflammaging at the molecular, cellular, organ, and disease levels, and review current aging models, the implications of cutting-edge single cell technologies, as well as anti-aging strategies. Since preventing and alleviating aging-related diseases and improving the overall quality of life are the ultimate goals of aging research, our review highlights the critical features and potential mechanisms of inflammation and aging, along with the latest developments and future directions in aging research, providing a theoretical foundation for novel and practical anti-aging strategies.

The massless solutions to the Dirac equation are described by the so-called Weyl Hamiltonian. The Weyl equation requires a particle to have linear dispersion in all three dimensions while being doubly degenerate at a single momentum point. These Weyl points are topological monopoles of quantized Berry flux exhibiting numerous unusual properties. We performed angle-resolved microwave transmission measurements through a double-gyroid photonic crystal with inversion-breaking where Weyl points have been theoretically predicted to occur. The excited bulk states show two linear dispersion bands touching at four isolated points in the three-dimensional Brillouin zone, indicating the observation of Weyl points. This work paves the way to a variety of photonic topological phenomena in three dimensions.

Large language models (LLMs), such as ChatGPT and GPT4, are making new waves in the field of natural language processing and artificial intelligence, due to their emergent ability and generalizability. However, LLMs are black-box models, which often fall short of capturing and accessing factual knowledge. In contrast, Knowledge Graphs (KGs), Wikipedia and Huapu for example, are structured knowledge models that explicitly store rich factual knowledge. KGs can enhance LLMs by providing external knowledge for inference and interpretability. Meanwhile, KGs are difficult to construct and evolve by nature, which challenges the existing methods in KGs to generate new facts and represent unseen knowledge. Therefore, it is complementary to unify LLMs and KGs together and simultaneously leverage their advantages. In this article, we present a forward-looking roadmap for the unification of LLMs and KGs. Our roadmap consists of three general frameworks, namely, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1) KG-enhanced LLMs,</i> which incorporate KGs during the pre-training and inference phases of LLMs, or for the purpose of enhancing understanding of the knowledge learned by LLMs; <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2) LLM-augmented KGs,</i> that leverage LLMs for different KG tasks such as embedding, completion, construction, graph-to-text generation, and question answering; and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3) Synergized LLMs + KGs</i> , in which LLMs and KGs play equal roles and work in a mutually beneficial way to enhance both LLMs and KGs for bidirectional reasoning driven by both data and knowledge. We review and summarize existing efforts within these three frameworks in our roadmap and pinpoint their future research directions.

Amyloid β protein (Aβ) is the main component of neuritic plaques in Alzheimer's disease (AD), and its accumulation has been considered as the molecular driver of Alzheimer's pathogenesis and progression. Aβ has been the prime target for the development of AD therapy. However, the repeated failures of Aβ-targeted clinical trials have cast considerable doubt on the amyloid cascade hypothesis and whether the development of Alzheimer's drug has followed the correct course. However, the recent successes of Aβ targeted trials have assuaged those doubts. In this review, we discussed the evolution of the amyloid cascade hypothesis over the last 30 years and summarized its application in Alzheimer's diagnosis and modification. In particular, we extensively discussed the pitfalls, promises and important unanswered questions regarding the current anti-Aβ therapy, as well as strategies for further study and development of more feasible Aβ-targeted approaches in the optimization of AD prevention and treatment.

Abstract M1 macrophage accumulation and excessive inflammation are commonly encountered issues in diabetic wounds and can fail in the healing process. Hence, hydrogel dressings with immunoregulatory capacity have great promise in the clinical practice of diabetic wound healing. However, current immunoregulatory hydrogels are always needed for complex interventions and high‐cost treatments, such as cytokines and cell therapies. In this study, a novel glycyrrhizic acid (GA)‐based hybrid hydrogel dressing with intrinsic immunoregulatory properties is developed to promote rapid diabetic wound healing. This hybrid hydrogel consists of interpenetrating polymer networks composed of inorganic Zn 2+ ‐induced self‐assembled GA and photo‐crosslinked methyl acrylated silk fibroin (SF), realizing both excellent injectability and mechanical strength. Notably, the SF/GA/Zn hybrid hydrogel can regulate macrophage responses in the inflammatory microenvironment, circumventing the use of any additives. The immunomodulatory properties of the hydrogel can be harnessed for safe and efficient therapeutics that accelerate the three phases of wound repair and serve as a promising dressing for the management of diabetic wounds.

Abstract This review discussed the dilemma of small data faced by materials machine learning. First, we analyzed the limitations brought by small data. Then, the workflow of materials machine learning has been introduced. Next, the methods of dealing with small data were introduced, including data extraction from publications, materials database construction, high-throughput computations and experiments from the data source level; modeling algorithms for small data and imbalanced learning from the algorithm level; active learning and transfer learning from the machine learning strategy level. Finally, the future directions for small data machine learning in materials science were proposed.

Communication systems to date primarily aim at reliably communicating bit sequences. Such an approach provides efficient engineering designs that are agnostic to the meanings of the messages or to the goal that the message exchange aims to achieve. Next generation systems, however, can be potentially enriched by folding message semantics and goals of communication into their design. Further, these systems can be made cognizant of the context in which communication exchange takes place, thereby providing avenues for novel design insights. This tutorial summarizes the efforts to date, starting from its early adaptations, semantic-aware and task-oriented communications, covering the foundations, algorithms and potential implementations. The focus is on approaches that utilize information theory to provide the foundations, as well as the significant role of learning in semantics and task-aware communications.

Session-based recommendation, which aims to predict the user's immediate next action based on anonymous sessions, is a key task in many online services (e.g., e-commerce, media streaming). Recently, Self-Attention Network (SAN) has achieved significant success in various sequence modeling tasks without using either recurrent or convolutional network. However, SAN lacks local dependencies that exist over adjacent items and limits its capacity for learning contextualized representations of items in sequences. In this paper, we propose a graph contextualized self-attention model (GC-SAN), which utilizes both graph neural network and self-attention mechanism, for session-based recommendation. In GC-SAN, we dynamically construct a graph structure for session sequences and capture rich local dependencies via graph neural network (GNN). Then each session learns long-range dependencies by applying the self-attention mechanism. Finally, each session is represented as a linear combination of the global preference and the current interest of that session. Extensive experiments on two real-world datasets show that GC-SAN outperforms state-of-the-art methods consistently.

Macrophages are critical mediators of tissue homeostasis, with the function of tissue development and repair, but also in defense against pathogens. Tumor-associated macrophages (TAMs) are considered as the main component in the tumor microenvironment and play an important role in tumor initiation, growth, invasion, and metastasis. Recently, metabolic studies have revealeded specific metabolic pathways in macrophages are tightly associated with their phenotype and function. Generally, pro-inflammatory macrophages (M1) rely mainly on glycolysis and exhibit impairment of the tricarboxylic acid (TCA) cycle and mitochondrial oxidative phosphorylation (OXPHOS), whereas anti-inflammatory macrophages (M2) are more dependent on mitochondrial OXPHOS. However, accumulating evidence suggests that macrophage metabolism is not as simple as previously thought. This review discusses recent advances in immunometabolism and describes how metabolism determines macrophage phenotype and function. In addition, we describe the metabolic characteristics of TAMs as well as their therapeutic implications. Finally, we discuss recent obstacles facing this area as well as promising directions for future study.

Time series prediction can be generalized as a process that extracts useful information from historical records and then determines future values. Learning long-range dependencies that are embedded in time series is often an obstacle for most algorithms, whereas LSTM solutions, as a specific kind of scheme in deep learning, promise to effectively overcome the problem. In this article, we first give a brief introduction to the structure and forward propagation mechanism of LSTM. Then, aiming at reducing the considerable computing cost of LSTM, we put forward a RCLSTM model by introducing stochastic connectivity to conventional LSTM neurons. Therefore, RCLSTM exhibits a certain level of sparsity and leads to a decrease in computational complexity. In the field of telecommunication networks, the prediction of traffic and user mobility could directly benefit from this improvement as we leverage a realistic dataset to show that for RCLSTM, the prediction performance comparable to LSTM is available, whereas considerably less computing time is required. We strongly argue that RCLSTM is more competent than LSTM in latency-stringent or power-constrained application scenarios.

Graph neural networks (GNN) has been considered as an attractive modelling method for molecular property prediction, and numerous studies have shown that GNN could yield more promising results than traditional descriptor-based methods. In this study, based on 11 public datasets covering various property endpoints, the predictive capacity and computational efficiency of the prediction models developed by eight machine learning (ML) algorithms, including four descriptor-based models (SVM, XGBoost, RF and DNN) and four graph-based models (GCN, GAT, MPNN and Attentive FP), were extensively tested and compared. The results demonstrate that on average the descriptor-based models outperform the graph-based models in terms of prediction accuracy and computational efficiency. SVM generally achieves the best predictions for the regression tasks. Both RF and XGBoost can achieve reliable predictions for the classification tasks, and some of the graph-based models, such as Attentive FP and GCN, can yield outstanding performance for a fraction of larger or multi-task datasets. In terms of computational cost, XGBoost and RF are the two most efficient algorithms and only need a few seconds to train a model even for a large dataset. The model interpretations by the SHAP method can effectively explore the established domain knowledge for the descriptor-based models. Finally, we explored use of these models for virtual screening (VS) towards HIV and demonstrated that different ML algorithms offer diverse VS profiles. All in all, we believe that the off-the-shelf descriptor-based models still can be directly employed to accurately predict various chemical endpoints with excellent computability and interpretability.

Ischemic stroke, which accounts for 75-80% of all strokes, is the predominant cause of morbidity and mortality worldwide. The post-stroke immune response has recently emerged as a new breakthrough target in the treatment strategy for ischemic stroke. Glial cells, including microglia, astrocytes, and oligodendrocytes, are the primary components of the peri-infarct environment in the central nervous system (CNS) and have been implicated in post-stroke immune regulation. However, increasing evidence suggests that glial cells exert beneficial and detrimental effects during ischemic stroke. Microglia, which survey CNS homeostasis and regulate innate immune responses, are rapidly activated after ischemic stroke. Activated microglia release inflammatory cytokines that induce neuronal tissue injury. By contrast, anti-inflammatory cytokines and neurotrophic factors secreted by alternatively activated microglia are beneficial for recovery after ischemic stroke. Astrocyte activation and reactive gliosis in ischemic stroke contribute to limiting brain injury and re-establishing CNS homeostasis. However, glial scarring hinders neuronal reconnection and extension. Neuroinflammation affects the demyelination and remyelination of oligodendrocytes. Myelin-associated antigens released from oligodendrocytes activate peripheral T cells, thereby resulting in the autoimmune response. Oligodendrocyte precursor cells, which can differentiate into oligodendrocytes, follow an ischemic stroke and may result in functional recovery. Herein, we discuss the mechanisms of post-stroke immune regulation mediated by glial cells and the interaction between glial cells and neurons. In addition, we describe the potential roles of various glial cells at different stages of ischemic stroke and discuss future intervention targets.

Interfacial layers (ILs) are prerequisites to form the selective charge transport for high-performance organic photovoltaics (OPVs) but mostly result in considerable parasitic absorption loss. Trimming the ILs down to a mono-molecular level via the self-assembled monolayer is an effective strategy to mitigate parasitic absorption loss. However, such a strategy suffers from inferior electrical contact with low surface coverage on rough surfaces and poor producibility. To address these issues, here, the self-assembled interlayer (SAI) strategy is developed, which involves a thin layer of 2-6 nm to form a full coverage on the substrate via both covalent and van der Waals bonds by using a self-assembled molecule of 2-(9H-carbazol-9-yl) (2PACz). Via the facile spin coating without further rinsing and annealing process, it not only optimizes the electrical and optical properties of OPVs, which enables a world-record efficiency of 20.17% (19.79% certified) but also simplifies the tedious processing procedure. Moreover, the SAI strategy is especially useful in improving the absorbing selectivity for semi-transparent OPVs, which enables a record light utilization efficiency of 5.34%. This work provides an effective strategy of SAI to optimize the optical and electrical properties of OPVs for high-performance and solar window applications.

To process and transfer large amounts of data in emerging wireless services, it has become increasingly appealing to exploit distributed data communication and learning. Specifically, edge learning (EL) enables local model training on geographically disperse edge nodes and minimizes the need for frequent data exchange. However, the current design of separating EL deployment and communication optimization does not yet reap the promised benefits of distributed signal processing, and sometimes suffers from excessive signalling overhead, long processing delay, and unstable learning convergence. In this paper, we provide an overview on practical distributed EL techniques and their interplay with advanced communication optimization designs. In particular, typical performance metrics for dual-functional learning and communication networks are discussed. Also, recent achievements of enabling techniques for the dual-functional design are surveyed with exemplifications from the mutual perspectives of “communications for learning” and “learning for communications.” The application of EL techniques within a variety of future communication systems are also envisioned for beyond 5G (B5G) wireless networks. For the application in goal-oriented semantic communication, we present a first mathematical model of the goal-oriented source entropy as an optimization problem. In addition, from the viewpoint of information theory, we identify fundamental open problems of characterizing rate regions for communication networks supporting distributed learning-and-computing tasks. We also present technical challenges as well as emerging application opportunities in this field, with the aim of inspiring future research and promoting widespread developments of EL in B5G.

TP53 is a critical tumor-suppressor gene that is mutated in more than half of all human cancers. Mutations in TP53 not only impair its antitumor activity, but also confer mutant p53 protein oncogenic properties. The p53-targeted therapy approach began with the identification of compounds capable of restoring/reactivating wild-type p53 functions or eliminating mutant p53. Treatments that directly target mutant p53 are extremely structure and drug-species-dependent. Due to the mutation of wild-type p53, multiple survival pathways that are normally maintained by wild-type p53 are disrupted, necessitating the activation of compensatory genes or pathways to promote cancer cell survival. Additionally, because the oncogenic functions of mutant p53 contribute to cancer proliferation and metastasis, targeting the signaling pathways altered by p53 mutation appears to be an attractive strategy. Synthetic lethality implies that while disruption of either gene alone is permissible among two genes with synthetic lethal interactions, complete disruption of both genes results in cell death. Thus, rather than directly targeting p53, exploiting mutant p53 synthetic lethal genes may provide additional therapeutic benefits. Additionally, research progress on the functions of noncoding RNAs has made it clear that disrupting noncoding RNA networks has a favorable antitumor effect, supporting the hypothesis that targeting noncoding RNAs may have potential synthetic lethal effects in cancers with p53 mutations. The purpose of this review is to discuss treatments for cancers with mutant p53 that focus on directly targeting mutant p53, restoring wild-type functions, and exploiting synthetic lethal interactions with mutant p53. Additionally, the possibility of noncoding RNAs acting as synthetic lethal targets for mutant p53 will be discussed.

Serious climate changes and energy-related environmental problems are currently critical issues in the world. In order to reduce carbon emissions and save our environment, renewable energy harvesting technologies will serve as a key solution in the near future. Among them, triboelectric nanogenerators (TENGs), which is one of the most promising mechanical energy harvesters by means of contact electrification phenomenon, are explosively developing due to abundant wasting mechanical energy sources and a number of superior advantages in a wide availability and selection of materials, relatively simple device configurations, and low-cost processing. Significant experimental and theoretical efforts have been achieved toward understanding fundamental behaviors and a wide range of demonstrations since its report in 2012. As a result, considerable technological advancement has been exhibited and it advances the timeline of achievement in the proposed roadmap. Now, the technology has reached the stage of prototype development with verification of performance beyond the lab scale environment toward its commercialization. In this review, distinguished authors in the world worked together to summarize the state of the art in theory, materials, devices, systems, circuits, and applications in TENG fields. The great research achievements of researchers in this field around the world over the past decade are expected to play a major role in coming to fruition of unexpectedly accelerated technological advances over the next decade.

Synaptic abnormalities are a cardinal feature of Alzheimer's disease (AD) that are known to arise as the disease progresses. A growing body of evidence suggests that pathological alterations to neuronal circuits and synapses may provide a mechanistic link between amyloid β (Aβ) and tau pathology and thus may serve as an obligatory relay of the cognitive impairment in AD. Brain-derived neurotrophic factors (BDNFs) play an important role in maintaining synaptic plasticity in learning and memory. Considering AD as a synaptic disorder, BDNF has attracted increasing attention as a potential diagnostic biomarker and a therapeutical molecule for AD. Although depletion of BDNF has been linked with Aβ accumulation, tau phosphorylation, neuroinflammation and neuronal apoptosis, the exact mechanisms underlying the effect of impaired BDNF signaling on AD are still unknown. Here, we present an overview of how BDNF genomic structure is connected to factors that regulate BDNF signaling. We then discuss the role of BDNF in AD and the potential of BDNF-targeting therapeutics for AD.

Polydopamine (PDA) nanoparticles have emerged as an attractive biomimetic photothermal agent in photothermal antibacterial therapy due to their ease of synthesis, good biodegradability, long-term safety, and excellent photostability. However, the therapeutic effects of PDA nanoparticles are generally limited by the low photothermal conversion efficiency (PCE). Herein, PDA@Ag nanoparticles are synthesized via growing Ag on the surface of PDA nanoparticles and then encapsulated into a cationic guar gum (CG) hydrogel network. The optimized CG/PDA@Ag platform exhibits a high PCE (38.2%), which is more than two times higher than that of pure PDA (16.6%). More importantly, the formulated CG/PDA@Ag hydrogel with many active groups can capture and kill bacteria through effective interactions between hydrogel and bacteria, thereby benefiting the antibacterial effect. As anticipated, the designed CG/PDA@Ag system combined the advantages of PDA@Ag nanoparticles (high PCE) and hydrogel (preventing aggregation of PDA@Ag nanoparticles and possessing inherent antibacterial ability) is demonstrated to have superior antibacterial efficacy both in vitro and in vivo. This study develops a facile approach to boost the PCE of PDA for photothermal antibacterial therapy, providing a significant step forward in advancing the application of PDA nano-photothermal agents.