Cystic echinococcosis caused by Echinococcus granulosus sensu lato remains a major One Health concern. In this study, we conducted a PRISMA-guided systematic review (Web of Science, 2005-2025) and compiled a georeferenced dataset of canine E. granulosus s.l. infections derived from published field studies. After data curation and harmonization, 169 location-year records (representing 159 unique spatial locations in SaTScan) were included in the space-time analysis. Space-time cluster detection using SaTScan (Bernoulli model; annual resolution) identified 14 clusters, of which 13 were statistically significant (P < 0.001). After excluding zero-radius clusters, seven regional clusters were retained for spatial interpretation, of which six were statistically significant. Major clusters were located in East/Central Asia, South America, Europe-North Africa, and the Middle East, indicating geographically structured and temporally bounded patterns of elevated infection risk rather than a homogeneous global distribution. In parallel, ecological niche modelling using Maxent (WorldClim bioclimatic variables and elevation; 2.5 arc-min resolution; 10 bootstrap replicates) demonstrated good model performance (training AUC = 0.914; test AUC = 0.910). Annual mean temperature (54.8% contribution) and elevation (23.5%) were the dominant predictors, with additional contributions from precipitation-related variables. Predicted suitability showed a unimodal response to climatic gradients, with highest suitability under moderate temperature and precipitation conditions. Using literature-derived records from published studies, our analyses identified spatial clusters and environmental variables associated with the distribution of canine E. granulosus s.l. These findings provide a broad, evidence-based risk assessment rather than a fully representative global surveillance picture. Overall, this integrated framework combining space-time clustering and ecological niche modelling provides a robust, evidence-based approach to identify priority areas for surveillance and targeted control of canine-mediated transmission.
Dynamic acoustofluidics enables precise, contact-free manipulation of particles, colloids, and cells and shows great potential for applications in physics, materials science, and life sciences. However, existing strategies struggle to realize contrast-based selective manipulation primarily because the pressure fields are time invariant. Here, we introduce a space-time acoustofluidic tweezer (STAT) that uses frequency detuning-induced pseudo-space-time modulation of standing surface acoustic waves to enable dynamic, contrast-dependent control of microparticles and cells. Experiments and simulations show that, under STAT manipulation, positive (PACP) and negative (NACP) acoustic contrast particles can undergo low-frequency, shear- and longitudinal-like harmonic motions, respectively. Under certain driving conditions, NACPs can be selectively guided along programmed paths, whereas PACPs remain stably patterned. Overall, STAT offers a gentle, biocompatible way to selectively drive oscillation, transport, and sorting among particles and cells of different acoustic contrasts, broadening the capabilities of acoustofluidic systems for biomedical applications.
In this paper, we present a conforming space-time discretization of the wave equation based on a first-order-in-time variational formulation. Our method extends the scheme of French and Peterson (1996), incorporating exponential weights in time, which yield an inf-sup stability condition for arbitrary choices of discrete subspaces, including spline spaces, without restrictions on the mesh size or time step. Moreover, using elliptic projections, we derive optimal convergence rates in both the energy and L 2 norms for sufficiently smooth solutions and for any choice of space-time tensor product subspaces satisfying standard approximation assumptions. Numerical examples are provided to support the theoretical findings.
Recent advances in high-dimensional multiplexing have enabled the concurrent operation of multiple independent communication channels through orbital angular momentum, polarization, and frequency division multiplexing, all implemented on a compact space-time-coding metasurface platform. These developments provide a streamlined and high-efficiency approach to optimizing multiplexing performance and enhancing channel capacity in wireless communication systems.
Atrial septal defect (ASD) has become increasingly common in the USA and now affects 1 in 11.3 children in some places, but space-time analysis has not been applied to this emerging trend. ASD rate (ASDR) data were obtained from the National Birth Defects Prevention Network 2003-2020. Substance (cigarettes, alcohol, cannabis, analgesics, cocaine) use data were obtained from the National Survey of Drug Use and Health. Income data were obtained from the US Census. Analysis was limited to the Non-Hispanic White population by technical factors. Time-sequential univariate and bivariate maps were prepared for both covariates and outcomes and their combinations. Spatial regression of the ASDR was performed using the R package splm. A total of 7.6% of data was interpolated by linear regression. A total of 110,107 ASD cases were identified amongst 17,751,437 live births in 27 US states across 10 reporting periods. Time series maps showed that ASDR showed concordant patterns with indices of cannabis use rather than other substances. This was confirmed by multivariate spatial regression where cannabis and cannabinoids alone were found to significantly relate to ASDR, with p = 0.00002 for cannabidiol. Cannabis legal status similarly tracked with ASDR. Compared to states where cannabis was not legal, ASDR was more prevalent in cannabis-legal states (OR = 2.73 (2.66, 2.80); E-Value 4.90 (lower C.I. 4.76)). Twenty-seven of 34 (79.4%) E-values were >9 (high range) and 34/34 were > 1.25 (causal threshold). Data show that cannabis, including cannabis legalization, is driving the US ASD epidemic. While most high-ASDR states have high rates of cannabis use, Midwestern states where cannabis is farmed, such as Kentucky, Tennessee and Missouri, do not, suggesting other routes of exposure, potentially implicating environmental contamination. ASD is a bellwether marker for cannabinoid teratogenicity, indicating that communities should carefully control cannabinoid exposure and limit transgenerational cannabinoid genotoxicity more generally.
Temporal metamaterials, created by modulating the refractive index in time, offer powerful means of controlling wave propagation but still lack a systematic design methodology. Here, we develop an analytic inverse-design framework rooted in space-time duality and the established theory of one-dimensional spatial inverse scattering. By prescribing reflection (backward-wave) and transmission (forward-wave) responses in rational-function form, we obtain closed-form refractive-index modulations that are guaranteed to be physically admissible. This approach avoids iterative optimization and provides direct analytic control of the modulation. We illustrate the method with syntheses of mathematical operators, such as derivatives and integrals, as well as Chebyshev- and Butterworth-type filters, and validate the results through finite-difference time-domain simulations. Our findings establish a general route to temporal media with tailored functional and spectral responses, enabling applications in wave-based information processing, programmable filtering, and amplification schemes inspired by photonic time crystals.
The neocortex covers a vast expanse of the mammalian brain and represents the principal target of clinical neuromodulation; however, the global principles of neocortical operation have been challenging to identify. In this regard, a limitation has been tracking activity with cortex-wide spatial coverage while maintaining access to the millisecond temporal resolution of neuronal firing- a crucial combination not achievable with existing recording technologies. Here we introduce and apply conformal immersion microscopy, enabling activity tracking across the entire dorsal cortex with millisecond temporal resolution and 100 μm spatial resolution (thus spanning five orders of magnitude in time and four in space), at sufficient sensitivity to resolve single-trial activity beyond 100 Hz. Drawing on physics-based frameworks, we apply multiscale analysis to identify a fundamental frequency-dependent coherence length that partitions the neocortex into discrete dynamical elements with well-defined propagation speeds, boundaries, and scale-invariant dynamics. These dynamical elements were found to be conserved from sub-threshold to suprathreshold (neuronal firing) regimes of neural activity, and were robust to diverse pharmacological, optogenetic, and genetic interventions. However, it was possible to identify and establish conditions allowing elemental boundaries to be selectively overridden, and to allow perturbation of specific elements even while conserving global dynamical architecture. Together, these findings enable measurement of intrinsic spatiotemporal parameters governing the dynamical organization of neocortex, which may provide a foundation for mechanistically-informed basic and translational understanding.
This study investigates transient heat conduction in a separable class of anisotropic functionally graded materials with spatially and temporally varying thermal properties and an internal heat source term. The governing transient heat equation with variable coefficients is first transformed into an equivalent form with constant coefficients through an appropriate variable transformation under prescribed separability conditions on the material properties. The resulting equation is then mapped into the Laplace domain to eliminate the time derivative and simplify the transient formulation. A fundamental solution in the Laplace space is employed to reformulate the transformed equation into a boundary integral representation containing source-related domain integrals, which serves as the basis for the boundary element formulation. The resulting integral equations are discretized using boundary elements, while the domain integrals are evaluated numerically through domain discretization in conjunction with particular solutions. The transient solution in the physical time domain is recovered by applying the Stehfest numerical inversion technique for the Laplace transform. Several benchmark test problems with known analytical solutions are presented to assess the accuracy, convergence, and stability of the proposed approach. Numerical results demonstrate that the developed boundary element formulation provides accurate and consistent solutions for transient heat conduction problems in anisotropic functionally graded materials satisfying the prescribed transformation conditions.
This study aimed to investigate the impact of spatial perception on time perception, and vice versa, in both auditory and visual modalities, as well as the influence of task difficulty in the visual modality. The experiment was conducted with children aged 5-6 years, 7-8 years, and adults. Participants completed four tasks over two days: one temporal bisection task and one spatial bisection task in each modality. In the visual modality, stimuli consisted of dots appearing on a screen, whereas in the auditory modality, spatially distributed sounds were presented. In both cases, participants were asked to indicate whether the duration or the distance was "short" or "long". Data analyses included ANOVAs, calculations of Weber Ratios (WR) and Bisection Points (BP). Results revealed significant cross-dimensional interference: space influenced time perception more strongly in the visual modality (kappa effect), while time further influenced space in the auditory modality (tau effect). These interference effects were more pronounced in older participants and modulated by task difficulty, with greater difficulty associated with reduced space-on-time interference (visual modality only). Children exhibited reduced interference despite noisier temporal representations, and the younger ones showed more symmetrical relationships, whereas adults displayed asymmetrical, modality-specific patterns. Overall, the findings highlight a strong developmental component in spatiotemporal interactions and support an Attentional-Bayesian framework that reconciles competing theories (ATOM and Conceptual Metaphor Theory) by positing a single flexible system where default Bayesian integration of magnitudes is dynamically modulated.
The many-body perturbation theory within the GW approximation is a widely used method for describing the electronic band structures in real materials. Its application to large-scale systems is, however, impeded by its high computational cost. The rate-limiting steps in a typical GW implementation are the evaluation of the polarization function under the random phase approximation (RPA) and the evaluation of the GW self-energy, both of which have a canonical O(N4) scaling with N being the system size. The conventional space-time algorithm within the plane-wave basis sets reduces the scaling from O(N4) to O(N3), albeit with a large prefactor and increased memory cost. Here, we present a space-time algorithm within the numerical atomic orbital (NAO) basis-set framework, for which the evaluation of the polarization function and self-energy is formally reduced to O(N2) or better with respect to system size. This is achieved by computing these quantities in real space, where low-scaling algorithms can be formulated by leveraging the localized resolution of identity (LRI) technique. The resulting NAO-based, LRI-enhanced space-time GW algorithm has been implemented in the LibRPA library interfaced with the FHI-aims code package. Benchmark calculations for crystalline solids show that the low-scaling implementation yields quasi-particle energies in close agreement with the conventional O(N4) k-space formalism previously implemented in FHI-aims. For the systems studied here, the observed overall scaling is substantially reduced relative to the canonical approach, and the low-scaling implementation becomes advantageous already for systems containing fewer than 100 atoms.
A model polyolefin mixturecomposed of 50% HDPE, 40% LDPE, and 10% PPhas been converted into valuable light olefins on an HZSM-5-based catalyst in a two-step continuous process: fast pyrolysis in a conical spouted bed reactor followed by upgrading in a catalytic fixed-bed reactor. The effect of the operating conditions in the catalytic step (space-time of 10 and 15 gcatalyst min gplastic -1 and temperature of 450 and 500 °C) on product yields and catalyst stability has been investigated up to a time on stream (TOS) of 4 h. Light olefins represented the most abundant product, with a maximum yield of 77% obtained at 500 °C with a space-time of 10 gcatalyst min gplastic -1 under zero-time conditions. Further increase in space-time led to a reduction in light olefin yield and to an increase of benzene, toluene, and xylene isomers (BTX) as well as light alkanes. Characterization of spent catalyst samples (coke deposition, physical properties) revealed the presence of two different types of coke and a higher coke deposition in the first section of the catalytic bed (in contact with incoming waxes). The catalyst proved to be stable at 500 °C, while higher deactivation was observed after 4 h of operation at the lower temperature (450 °C).
2'-Deoxy-2'-fluoroadenosine (2'-F-dA) is a nucleoside analogue used as a key building block for oligonucleotide drugs. It can be biosynthesized from a low-cost 2'-deoxy-2'-fluorouridine via one-pot transglycosylation catalyzed by a thymidine phosphorylase (TP) and a purine nucleoside phosphorylase (PNP). However, reliance on purified enzymes and low space-time yields present challenges for industrial application of the process. Here, we develop a whole-cell-based biocatalytic system employing TP and PNP from Escherichia coli, which demonstrates high catalytic efficiency and operational simplicity in scaled-up reaction. In particular, a thermal pretreatment of TP- and PNP-expressing whole cells, determined as 50 °C for 3 h, effectively suppressed endogenous deamination side reaction while enhancing 2'-F-dA yield. Subsequent optimization of enzyme and substrate loadings and their relative ratios achieved an unprecedented space-time yield of 1.22 g/L/h with 88.1 g/L product titer in a 500 mL scaled-up reaction, manifesting a highest total conversion of 68.2 %. An integrated purification process yielded gram-scale solid powder of 2'-F-dA with 98.0 % chemical purity and 85.0 % recovery. This novel whole-cell biocatalytic process demonstrates significant industrial potential for the production of 2'-F-dA.
Sierra Leone, using data from the District Health Information System 2 (DHIS2) database. This study examined the spatial and spatio-temporal distribution of malaria incidence, and the relationship between malaria incidence and rainfall in surrounding areas in Sierra Leone from 2021 to 2024. A cross-sectional geospatial study was conducted using malaria case data, mean rainfall data, population estimates, and chiefdom-level geographic coordinates. Spatial clustering was evaluated using Moran's I, significant district-level clusters were identified through space-time Poisson models (α = 0.05; 999 permutations), and the relationship between malaria incidence and rainfall in surrounding areas was assessed using bivariate Moran's I, implemented in Python. Between 2021 and 2024, 7.4 million malaria cases were reported. Chiefdom-level incidence ranged from 21.2 to >750 per 1,000 population. Significant spatial clustering was observed (Moran's I > 0; P < 0.01). Persistent high-high clusters (P < 0.05) and low-low clusters (P < 0.05) were identified across the country. Space-time analysis identified both high-risk (relative risk [RR] = 1.2-1.9; P < 0.01) and low-risk districts (RR = 0.5-0.9; P < 0.01). There was no significant association between malaria incidence and rainfall in surrounding areas at the national level. Malaria transmission remains spatially and temporally heterogeneous, with persistent hotspots that require tailored subnational interventions to accelerate progress towards elimination.
Our study investigated metaphor inferencing in second languages (L2). According to universalist accounts, L2 metaphor comprehension should be intuitive. However, cross-linguistic differences suggest that successful metaphor inferencing is subject to transfer effects. Little is known about the interaction of transfer effects involving modality differences in the context of sign languages. We investigated whether existing sign language expertise enhances SPACE-TIME metaphor inferencing in signed L2s. Crucially, we focused on SPACE-TIME metaphor inferencing because the need to abstract from salient concrete SPACE interpretations may pose challenges to sign novices. Methodologically, we compared native signers (N = 27, L1 = American or British Sign Language) to hearing sign novices (N = 37, L1 = English). Participants were exposed to a novel sign language through a short weather forecast, mirroring real-life implicit immersion. Subsequently, they completed tasks tapping into meaning assignment. Our analyses focused on metaphor inferencing, considering transfer effects from cognateness regarding L1-L2 sign similarity and gesture-sign similarity. Our findings revealed a marked dissociation in transfer effects by group. Amongst native signers metaphoricity facilitated L2 meaning assignment. However, metaphoricity posed a challenge to sign novices who defaulted to concrete SPACE interpretations, misled by their saliency and iconicity. At the item-level, cognate effects occurred both from existing linguistic knowledge and gestural repertoires. Our study suggests that whilst all participants recognised universal iconic patterns facilitating concrete meaning assignment, abstraction and metaphor inferencing relied on positive transfer from specific cognates or generic knowledge of modality-specific metaphor formation strategies. Hence, teaching materials should build on existing knowledge in a bespoke manner.
In the context of global climate change, the urban heat island effect (UHI) poses a severe threat to the health and life quality of residents. Explorations of the spatiotemporal variations of intra-urban heat island effect present as a key foundation for identifying the impacts of diversified and differential construction modes on the thermal environment, as well as testing the implementation effect of relevant planning interventions. However, the spatio-temporal variability of UHI manifests as highly dynamic and stochastic, governed by the nonlinear coupling between built environment and climate system. Such intrinsic volatility poses a critical challenge for precise diagnostic assessment in urban planning. To distill actionable insights from this complexity, two classes of "sequential reduction" strategies have emerged. The "time-space" approach prioritizes temporal trend extraction before spatial partitioning based on trend similarity, while the "space-time" approach clusters spatial units by trajectory similarity prior to trend analysis. Yet, despite their proliferation, a systematic critique regarding their core rationales, methodological protocols, and distinct domains of applicability remains conspicuously absent. To fill this gap, we utilized 11 periods of land surface temperature data by Landsat satellites for Wuhan from 2000 to 2024 to systematically compare the two methods in terms of analytical capabilities and applicability in planning evaluations. Results showed that the "time-space" method, relying on the trend test of independent pixels, identified 10 types of spatial zones, which could be further summarized into three major patterns, including persistent warming pattern, significantly mitigated pattern, and balanced-stable pattern. Among them, persistent warming pattern accounted for 21.9% of the area, while significant mitigation pattern accounted for 0.1%. Those results indicated that this method was highly sensitive to local subtle changes and was therefore more suitable for meso-micro assessment oriented to urban detailed planning and urban renewal. The "space-time" method was based on temporal clustering to preferentially identify homogeneous spatial units. It identified 13 types of spatial zones with consistent upward trends, which could be further summarized into three major patterns, including persistent warming pattern, stable high-temperature pattern, and stable low-temperature pattern. Among them, persistent warming pattern accounted for 26.6% of the area. This method emphasized the identification of the overall continuous spatial pattern and was hence more suitable for city-level assessment through master plans for improved spatial patterns of urban thermal environments. Overall, both categories of methods revealed a shared pattern in the spatiotemporal variations of the UHI in Wuhan's metropolitan development area: "overall intensification and core shift". However, distinct differences in the logic and strategies used to handle spatiotemporal complexity had led to significant discrepancies in the assessment results, indicating an urgent need for further methodological innovation. 全球气候变化背景下,城市热岛效应给居民健康及生活品质带来严峻威胁。城市内部热岛效应时空演变模式挖掘,是识别多元化和差异性建设模式对热环境影响、检验规划调控实施效果的关键基础。然而,热岛效应时空变化在人居环境与气候系统交互作用下呈现出高度动态性和不确定性特征,给规划精准评估带来巨大挑战。尽管近年来涌现了对时间或空间变化复杂性进行先后压缩、以提炼典型时空特征服务规划实践需求的两类方法(包括先提取时间变化趋势、再基于时间变化趋势的相似性进行空间分区的“时间-空间”法,以及先基于时间变化轨迹相似性进行空间分区、再提取时序趋势的“空间-时间”法),但目前学界对其核心思路、实施路径及适用场景仍缺乏系统讨论。对此,本研究基于武汉市2000—2024年11期Landsat地表温度数据,对比了上述两类方法的解析能力和规划适用性。结果表明:“时间-空间”法依托独立像元趋势检验,识别出10类空间,可归纳为持续增温、显著改善与均衡稳定3大类型。其中,持续增温区占比21.9%,显著改善区占比0.1%。该方法对局部非平稳突变敏感,适用于详细规划和城市更新等精细评估。“空间-时间”法基于时序聚类优先识别均质空间单元,得到13类空间,可归纳为持续增温、稳定高温与稳定低温3大类型。其中,持续增温区占比26.6%。该方法擅长捕捉整体连续空间格局,更适用于面向总体规划空间格局优化需求的城市层面整体评估。整体上,两类方法揭示武汉都市发展区热岛效应时空演变呈现“整体增强、核心转移”的共性,但不同方法对时间和空间复杂性处理逻辑与方式的区别仍导致了评估结果的显著差异,未来亟需进一步的方法创新。.
Microfluidic reaction formats are of increasing importance to produce low molecular weight chemicals. In-situ product recovery in such systems benefits from intrinsic physical advantages of microfluidics, including increased mass and energy transfer and continuous processing. This opinion highlights the design space for microfluidic in-situ recovery of gaseous products with respect to rates of production and product recovery, as well as physical constraints limiting product titers. In a theoretical case study, O2 produced by cyanobacteria in a microfluidic reactor is used to benchmark productivities through process simulations. In general, the productivity of microfluidic bioprocesses is confined by biocatalytic reaction rates, diffusion rates, and residence time. Key challenges for scale-up, and thus achieving high space-time yields, comprise biocatalyst formats, catalytic robustness, and specific turn-over rates, as well as numbering-up of microfluidic devices. Economically and physically accessible application areas may thus be limited to milligram or higher gram quantities of specialty products.
Spatiotemporal epidemiology of dengue remains poorly understood in the Philippines and there is scarcity of a nationwide spatiotemporal cluster analysis. This study utilizes long-term nationwide data to identify the spatial patterns and spatiotemporal clustering of dengue incidence in the Philippines. We obtained monthly data from January 2017 to December 2024 across all provinces from the Philippine Epidemiology Bureau. The data were analyzed via spatial analysis techniques, specifically Moran's I and local Getis-Ord Gi* to determine spatial autocorrelation and hotspots. Furthermore, Poisson and space-time permutation (STP) models with varying maximum reported cluster size (MRCS) settings were applied to identify dengue spatiotemporal clusters. A total of 1,903,425 dengue cases were reported in the study period, with a high concentration of cases consistently observed in the National Capital Region (NCR). Significant positive spatial autocorrelation was observed in the study period with hotspots varying across the years. Ifugao, Kalinga, Abra, Isabela and Mountain Province are the provinces most frequently identified as hotspots. Areas within the Western Visayas region were consistently identified under the primary clusters by the spatiotemporal models signifying the impact of the 2019 epidemic in the region. Compared with the Poisson models, the STP model had identified more clusters with smaller radii. To our knowledge, this is the first spatiotemporal cluster analysis in the Philippines on reported dengue cases at the national scale using spatial scan statistics. The study demonstrated the application of varying MRCS which has effectively detected meaningful clusters. These findings offer health agencies and authorities in the Philippines approaches to further understand disease epidemiology, particularly in terms of spatial and spatiotemporal clustering, which consequently enables the implementation of targeted interventions and resource allocation.
The catalytic valorization of CO2 and glycerol into glycerol carbonate (GC) provides a sustainable pathway for carbon management and biodiesel byproduct utilization. Herein, a series of In-doped ZnAl hydrotalcite-derived catalysts were developed for the dehydration-assisted synthesis of GC using 2-cyanopyridine. Structural characterization indicates that indium incorporation finely modulates the electronic environment of Zn and Al species, optimizing the distribution of surface Lewis acid-base sites. Moderate doping (x = 0.1) significantly enhanced the density of medium-strength basic sites crucial for CO2 activation while suppressing the strong basicity responsible for nonselective glycerol polymerization. Under 170 °C and 4.0 MPa of CO2, the optimal In0.1/Zn4Al-CHT catalyst exhibited a glycerol conversion of 70.2% and a GC selectivity of 68.8%, delivering a superior space-time yield (STY) of 6.71 mmol gcat-1 h-1. Mechanistic analysis suggests that the synergistic interaction between In-induced basic sites and Lewis acid centers facilitates the selective carbonylation pathway. Furthermore, the catalyst demonstrated robust structural stability and reusability over multiple reaction cycles without significant metal migration or sintering. These findings highlight the potential of indium-doped mixed oxides as efficient, non-noble bifunctional catalysts for industrial CO2 utilization.
A sustainable photocatalytic protocol for the trifluoromethylation of (hetero)arenes is reported. The method operates under metal- and base-free conditions using an inexpensive and atom-efficient CF3 source, trifluoroacetic anhydride, in ethyl acetate as a green solvent. An organic cyanoarene photocatalyst enables efficient CF3 radical generation under blue-light irradiation, providing a broad range of trifluoromethylated arenes and heteroarenes. The reaction displays pronounced sensitivity to substituent patterns rather than electronic effects. Mechanistic investigations, including radical trapping, Stern-Volmer analysis, and DFT calculations, support a reductive quenching pathway involving photocatalyst-mediated reduction of TFAA. The protocol is amenable to scale-up, with continuous-flow operation delivering a significant increase in space-time yield, highlighting its potential for sustainable synthesis of CF3-containing molecules.
Raspberry ketone synthase is the key enzyme for raspberry ketone biosynthesis, yet its activity and thermostability remain suboptimal for industrial use. By combining computational and deep-learning-guided design, we obtained a Rubus idaeus raspberry ketone synthase (RiRZS1) mutant M3 with 12.06-fold higher relative activity and 38.71-fold improved thermostability compared with the wild type. Kinetic analysis showed that M3 exhibited 14.89-fold higher kcat and 5.91-fold higher catalytic efficiency(kcat/Km), indicating markedly enhanced catalytic competence. Structural inspection and molecular dynamics simulations revealed that M3 induced only subtle conformational rearrangements but reduced global flexibility, accounting for its improved thermostability. Meanwhile, tunnel and substrate-pocket reshaping in M3 was associated with increased prereaction-state populations, which may contribute to enhanced catalytic performance. In whole-cell biotransformation, the raspberry ketone titer of M3 was 2.38-fold that of wild-type, and the corresponding space-time yield was 3.57-fold higher. This computational and artificial intelligence-assisted framework enabled a synergistic optimization of RiRZS1 activity and stability, and can serve as a transferable design framework for protein engineering in sustainable bio-manufacturing.