Hong Kong University of Science and Technology
Hong Kong University of Science and Technology Institutional RepositoryNot a member yet
162821 research outputs found
Sort by
From biodiversity to clinical translation: a global review of lichen-derived natural products and their pharmacological potential
No full textLichens, which are unique ecosystems formed by fungi and photosynthetic partners like algae or cyanobacteria, hold great potential for drug discovery. This review compiled four decades of research to create the first comprehensive global database of 14,230 lichen species, among which 116 were highlighted for proven medicinal properties. Our spatial analysis identified key regions for therapeutic potential, such as Europe, coastal North America, East Asia, and Oceania, with the Parmeliaceae family was found to contain the most bioactive species. We systematically described 82 lichen-derived metabolites with confirmed pharmacological effects. Depsidones show promise as agents against oxidative stress and cancer; depsides/polysaccharides demonstrate efficacy in modulating neuroinflammation and immune response; and terpenoids contribute to addressing antimicrobial resistance. Cytotoxic, anticancer, anti-tumor, anti-microbial, and antioxidant activities are associated with the largest number of medicinal species. This research bridges traditional knowledge of lichens with modern pharmacology, identifying 195 medicinal species that need further validation for clinical use. We also discussed challenges in translating this knowledge into practice, such as ensuring sustainable biomass, optimizing bioavailability, and ensuring clinical safety, and we suggested lichen-specific guidelines for these processes. As lichen metabolites offer compelling opportunities to tackle pressing global health issues, such as antimicrobial resistance, neurodegenerative disorders, and cancer, we advocate for lichen metabolites as promising solutions for future pharmaceutical development.</p
Daidzein-engineered high-accumulation self-therapeutic nanoplatform drives androgen receptor degradation via ubiquitination-proteasome system to overcome enzalutamide resistance in castration-resistant prostate cancer
No full textHigh levels of androgen receptor (AR) are associated with poor prognosis and drug resistance in castration-resistant prostate cancer (CRPC), and monotherapy with AR antagonists such as enzalutamide (Enz) is currently inadequate to meet the clinical needs. Dual-targeted therapeutic strategy of AR degradation combined with AR inhibition is a promising strategy to improve the efficacy and to address the problem of drug resistance in CRPC, thus the development of safer, stable, and effective agents with AR-degrading activity is urgent. In our study, we found that daidzein, a phytoestrogen, could reduce the risk of prostate cancer, exhibited a good affinity for AR and down-regulated AR levels. Consequently, we developed daidzein-based, redox-responsive self-therapeutic nanocarriers to deliver Enz (D44DA@Enz NPs) to realize the dual-targeted therapeutic strategy against AR in CRPC. The results showed that D44DA NPs facilitated AR degradation via the ubiquitin-proteasome pathway, which combined with the inhibitory effect of Enz on AR nuclear translocation greatly improved the efficacy against CRPC. In vivo studies, D44DA NPs demonstrated approximately 40-fold increased enrichment at the tumor site, exhibited excellent self-antitumor activity and synergistically enhanced the therapeutic effects of Enz with minimal toxic side effects. In summary, this study provides a highly promising AR-degrading active nanoplatform and dual-targeted AR therapeutic strategy for the treatment of CRPC.</p
The influence of cabin environment on takeover performance in conditional automated driving
No full textEnvironmental factors such as temperature and carbon dioxide concentration have been widely studied for their effects on cognitive and physical performance. However, their impact on takeover performance in the context of driving automation remains underexplored. This study investigated how temperature and carbon dioxide concentration influence drivers’ takeover performance in conditional automated driving. Using a driving simulator, the experimental setup simulated modern vehicle cabin conditions, with three levels of temperatures (i.e., slightly cool, neutral and slightly warm) and two realistic carbon dioxide concentrations (i.e., high level, corresponding to recirculation ventilation and low level, corresponding to outside air ventilation). In total, 60 gender-balanced drivers participated in the study and a between-subjects experiment design (temperature and carbon dioxide concentration) was adopted. Each participant experienced three types of typical takeover scenarios, as initiated by visual and auditory takeover requests in a vehicle with conditional driving automation. The results showed that a slightly cool temperature and high carbon dioxide concentration negatively affected longitudinal speed control, while a slightly warm temperature resulted in quicker takeover actions and more stable lateral control. The analyses of gaze-related metrics suggested that elevated carbon dioxide concentrations may increase fatigue, while slightly cool temperatures appeared to reduce fatigue, leading to faster on-road attention. Finally, high carbon dioxide can counteract the favorable effects of temperature at the cognitive level. This study provides insights into the design of cabin environment control to improve driving safety in the context of driving automation.</p
Soluble CSF1R alleviates microgliopathy in a CSF1R-related leukoencephalopathy (CRL) mouse model
No full textColony-stimulating factor 1 receptor (CSF1R), primarily expressed on microglia in the central nervous system (CNS), is essential for microglial homeostasis and survival. CSF1R dysfunction, due to a monoallelic mutation, causes CSF1R-related leukoencephalopathy (CRL), a primary microgliopathy. CSF1R undergoes proteolytic cleavage to release soluble CSF1R (sCSF1R), which is decreased in the serum of CRL patients. However, the biological function of sCSF1R remains unknown. Here, we found that sCSF1R alleviated cognitive impairment and anxiety-like behavior in Csf1r+/− mice. Importantly, we identified CSF1R as a target binding protein of sCSF1R on microglia. Notably, sCSF1R inhibited the activation and inflammatory factor expression of Csf1r+/− microglia by reducing the phosphorylation of CSF1R (Y723) and NF-κB (S468 and S536). These results demonstrate that sCSF1R exerts neuroprotective effects by binding membrane-bound CSF1R and inhibiting pathological microglial activation by inhibiting the nuclear translocation of NF-κB. These findings identify sCSF1R as a potential therapeutic agent for CRL.</p
Offline inverse reinforcement learning for joint optimization of energy costs and demand charge in industrial PV-battery load systems
No full textIndustrial electricity bills are typically composed of two major components: the energy charge, which is based on the total accumulated energy consumption over a billing period (e.g., one month), and the demand charge, which depends on the highest peak power observed during the same period. Consequently, the joint optimization of energy costs (through energy arbitrage) and demand charges (through peak shaving) is crucial for effective cost management in industrial PV-battery load systems. However, this task remains fundamentally challenging due to the volatility of renewable generation and load, the complex temporal dependencies introduced by peak demand charges, and the competing objectives between immediate cost savings and long-term peak reduction—rendering existing model-based and data-driven energy management approaches inadequate for real-world applications. To tackle these challenges, this paper formulates the problem as a soft Markov Decision Process (MDP) and proposes a novel Offline Inverse Reinforcement Learning (OIRL) framework based on a dual reward-policy iterative optimization mechanism. Our approach introduces an innovative synthesis of contrastive reward learning—leveraging both expert demonstrations and on-policy trajectory rollouts—with conservative soft Q-learning optimization. This architecture enables accurate reconstruction of implicit reward structures through comparative analysis of expert and agent behaviors, while ensuring stable policy improvement via regularized value function updates with pessimistic value initialization. Extensive experiments using real-world data from our industrial partner in China demonstrate that OIRL achieves substantial energy arbitrage and peak shaving improvement compared to state-of-the-art reinforcement learning baselines in energy management. Furthermore, the framework maintains robust performance across diverse operating conditions, establishing a new paradigm for intelligent control of industrial PV-battery load systems.</p
Amino acid-modified degradable environmentally friendly antifouling coatings loaded with natural antifoulant
No full textA novel eco-friendly marine antifouling coating was developed by synthesizing amino acid-modified polycaprolactone-based polyurethane. L-lysine, L-glutamic acid, and L-tyrosine served as initiators for the ring-opening polymerization of caprolactone, in combination with the biodegradable antifouling agent 5-octylfuran-2(5H)-one (butenolide). The modified polymers exhibited enhanced mechanical properties, increased hydrophilicity, and accelerated degradation rates, thereby facilitating the controlled release of the butenolide. The glutamic acid-modified polymer showed superior mechanical strength and adhesion. The tyrosine-modified polymer exhibited high hydrophilicity with fast degradation and release rates due to the rigid phenyl side chain. Meanwhile, the lysine-modified polymer achieved balanced strength and degradability. The anti-biofouling tests confirmed that all modified polymers demonstrated excellent performance. This study presents an effective molecular design strategy for developing green, biodegradable marine antifouling coatings and advances the practical application of butenolide-based antifouling systems.</p
Gear-tuning meta-shaft for low-frequency torsional vibration suppression
No full textMetastructures can be engineered with low-frequency torsional band gaps, which provides a new solution for vibration suppression in shaft systems. However, achieving precise, reversible and robust tunability remains challenging, particularly in shaft systems due to the limited space preventing the implementation of tunable designs with complex mechanisms or additional control units. In this study, a tunable meta-shaft with self-locking gear (SLG) resonators is proposed, where the vibration suppression frequency range of the meta-shaft can be adjusted precisely with a simple gear-tuning mechanism without adding resonator's mass. By shifting the SLG teeth to control the deformation of the six curved beams in the SLG resonators, the torsional stiffness and resonant frequency can be effectively modulated, thereby enabling the generation of tunable low-frequency torsional band gaps. The torsional wave attenuation performance of the meta-shaft with periodically attached SLG resonators is evaluated numerically, and a prototype is fabricated to experimentally verify its robust and tunable vibration suppression capability. Consistent results among theoretical analysis, numerical simulations, and experiments verify the effectiveness and scalability of the proposed tuning strategy for adaptable torsional vibration control in shaft systems.</p
A composite mean temperature transformation for compressible turbulent boundary layers
No full textIn the present study, we introduce a new temperature transformation for compressible turbulent boundary layers with adiabatic and isothermal walls. Unlike existing transformations that rely on a single invariant function for the non-dimensional temperature gradient across the entire inner layer, a composite transformation strategy is proposed by leveraging two newly proposed Mach-number and wall-temperature invariant functions for the mean temperature field. This approach not only deploys appropriate Mach-number invariant functions in the viscous sublayer and the logarithmic region, but also introduces an improved solution to the long-standing singularity challenge inherent in single invariant function models. The performance of this composite transformation is verified by extensive direct numerical simulation (DNS) datasets (26 cases) of compressible turbulent boundary-layer flows. The results demonstrate that the proposed transformation maps the mean temperature profiles to the incompressible reference without case-specific parameter tuning, exhibiting significantly reduced scatter when compared with the existing temperature transformations
A multi-domain machine learning framework for intelligent condition monitoring of marine diesel engines
No full textAccurate and early monitoring of injector faults in marine diesel engines is essential for maintaining operational efficiency and preventing unplanned downtime. This study develops a novel multi-domain machine learning framework based on wavelet transform (WT), principal component analysis (PCA), and dynamic independent component analysis (DICA). The framework extracts complementary statistical and spectral indicators from vibration signals, encompassing time, frequency, and time–frequency-domain representations. The multi-domain WT-based PCA–DICA framework ensures effective dimensionality reduction and separation of correlated sources, enabling the identification of hidden features that are difficult to capture through single-domain analysis. A diverse set of machine learning classifiers, including support vector machines, naïve Bayes, k-nearest neighbors, logistic regression, AdaBoost, and linear discriminant analysis, is systematically evaluated using a five-fold cross-validation approach. The proposed approach, tested and validated on the Sulzer 6AL20/24 test engine, achieves classification accuracies and AUC values exceeding 98%, demonstrating robustness to noise, class imbalance, and limited fault data. The proposed framework demonstrates robustness against noisy and unbalanced datasets through data augmentation techniques (slicing and random shuffling) and systematic feature concentration analysis. This multi-domain, data-driven framework provides a reliable, scalable approach to injector fault diagnosis and can be extended to other types of faults and engine configurations. The results provide valuable insights into condition-based maintenance and intelligent marine-engine health-monitoring systems.<br/
政策终结的比较案例研究——一个新的类型学框架
No full text政策终结是政策过程中的重要一环，通常体现为政策主体有意识地决策，并呈现出一定的形态规律。在公共政策过程理论中，关于政策终结类型及其机制的知识相对缺乏，限制了对现实世界多样政策终结实践的解释力。聚焦公共政策解决社会问题的一般规律，从“政策目标达成与否”与“政策终结完全与否”两个核心维度，抽象归纳出政策终结的五种类型：目标实现型、战略升级型、结构纠偏型、工具优化型和目标退出型。在此基础上，结合中国情境中的五个典型案例，进一步阐释上述政策终结的类型学框架。对政策终结现象的分析和归纳，有助于丰富和完善政策过程理论体系。Policy termination is an important phrase in the policy process, typically manifested as deliberate decisions by policymakers and demonstrating certain patterns． In public policy process theory, knowledge regarding the types and mechanisms of policy termination remains relatively under developed, limiting its explanatory pow er for the diverse practices of policy termination in the real world． By focusing on the general principles of how public policies address social issues, based on two core dimensions— “whether policy goals are achieved” and “whether termination is complete ”， this study abstractly summarize five types of policy termination: “mission accomplished” “strategic upgrading” “structural recalibration” “instrumental optimization” and “objective cessation”. Building on this framework, the study further elucidates this typology by analyzing five typical cases within the Chinese context． The analysis and categorization of policy termination phenomena contribute to enriching and refining the theoretical system of the policy process