Abstract: A series of critical interactions within the viral core between the viral RNA (vRNA) and HIV-1 Integrase (IN) has previously been reported. In these studies, contact points between vRNA and IN were identified using RNA-seq and MS-based protein foot-printing approaches. Several IN amino acids located in its C-terminal domain (CTD) were found to be essential for vRNA binding and their alanine substitution severely impacted the correct morphogenesis of the matured viral core. Here, we have extended these studies by performing a comprehensive mapping of the IN-vRNA interaction by deploying RNA crosslinking and NMR methodologies. Together, these approaches were able to identify additional contacts points between the vRNA and IN. Our results reveal several new basic amino acids located in the IN CTD critical for the vRNA-IN interaction, viral replication and correct morphology of the matured viral core.

Abstract: Oleanolic acid (OA) is a pentacyclic triterpenoid with broad biological activity, but its primary molecular points of engagement remain incompletely resolved. Most available studies describe OA through selected pathway markers, particularly within PI3K/AKT/mTOR, AMPK/mTOR, MAPK, NF-κB, and Nrf2 signaling, without clearly distinguishing direct target engagement from downstream adaptive responses. This limits mechanistic interpretation and weakens translational prioritization. This review examines how phosphoproteomics and integrated multi-omics can support OA target deconvolution. We discuss why phosphoproteomics is particularly informative for capturing early signaling events, how it can be combined with proteomics, tran-scriptomics, metabolomics, and chemoproteomic approaches, and why orthogonal tar-get-engagement methods remain essential for stronger causal inference. We also organize the current signaling evidence for OA and its derivatives, highlighting the strongest support for AMPK/mTOR-linked regulation of autophagy and apoptosis while identi-fying major gaps in systems-level validation across other reported pathways. Finally, we propose a stepwise workflow for OA target deconvolution based on time-resolved phosphoproteomics, analysis of informative phosphosite subsets, mul-ti-omics integration, kinase/phosphatase activity inference, and experimental target validation. This framework may help move OA research from descriptive pathway pharmacology toward mechanism-based target prioritization and more rational deriva-tive development.

Abstract: Objective: To describe a rare case of subglottic foreign body (FB) impaction in an adult, leading to subglottic stenosis, and to evaluate the effectiveness of Suspension Microlaryn-goscopy (SML) as a minimally invasive technique for its management. Methods: A single case of subglottic FB in an adult was managed using endotracheal intubation and SML. This approach allowed for precise visualization and extraction of the FB. When granula-tion tissue or mucosal overgrowth was encountered, targeted resection and controlled cauterization were performed. Results: The SML technique provided excellent exposure to the subglottic region, enabling safe removal of the FB while minimizing trauma to the surrounding tissues. Postoperative follow-up demonstrated successful resolution of sub-glottic stenosis without complications, confirming the efficacy and safety of this approach. Conclusions: Subglottic FB impaction is an exceptionally rare occurrence in adults, posing diagnostic and therapeutic challenges. SML is a valuable tool for managing such cases, offering precise visualization, safe FB extraction, and effective management of complica-tions such as granulation tissue, ensuring optimal patient outcomes.

Abstract: Binary Waring decomposition seeks to express a homogeneous binary form as a minimal sum of powers of linear forms. In the binary setting, Sylvester’s theorem gives a classical algebraic route for rank determination and parameter recovery through structured Hankel/catalecticant matrices. Although this procedure is exact and interpretable in ideal arithmetic, practical rank identification may become unstable when the input coefficients are contaminated by noise or when the underlying roots are close to degenerate configurations. This paper develops a data-driven rank inference framework coupled with certified Sylvester reconstruction for robust binary Waring decomposition. The proposed method first converts the coefficient sequence into a Hankel-aware graph that captures recurrence-induced dependencies among polynomial coefficients. A graph neural network is then used to infer plausible rank candidates from this structured representation. Instead of accepting a single prediction directly, the framework performs explicit Sylvester reconstruction and algebraic residual verification for candidate ranks. To further improve decision reliability, a lightweight meta-verification module integrates reconstruction residuals, model confidence scores, and stability-related indicators to select the most credible rank. Experiments on large-scale synthetic binary forms demonstrate that the proposed approach improves rank identification accuracy and verified reconstruction success under low-to-moderate noise, while maintaining the transparency and auditability of classical symbolic–numeric computation. These results suggest that data-driven rank inference can serve as an effective front-end for algebraically certified reconstruction, especially in numerically ambiguous regimes where fixed threshold-based Sylvester implementations are fragile.

Abstract: The high energy consumption of conventional mixers equipped with active mixing elements necessitates the development of more efficient technologies for mixing bulk materials and feed mixtures. This study proposes a gravity-based mixing method based on the rotation of an inclined cylindrical chamber without the use of active mixing elements. During rotation, the mixture components move toward both end walls of the chamber and simultaneously perform circular motion along the inner cylindrical surface, which intensifies the mixing process and reduces energy consumption. A structural and technological design of the gravity mixer was developed, and an experimental prototype was manufactured. Analytical relationships were obtained to determine the critical rotational speed of the chamber, particle movement velocity, and the power required for the mixing process. Laboratory experiments showed that the average particle movement velocity was 1.21 m/s and the average friction coefficient was 0.40. Under optimal operating conditions, the mixture uniformity reached 95.7% after 4 min of mixing. The mixer productivity was 0.95 t/h, while the specific energy consumption was 0.5 kWh/t, which is 2.5 times lower than that of conventional mixers equipped with active mixing elements. The obtained results confirm the potential of the proposed gravity-based mixing method for preparing feed and organomineral mixtures in small-scale farming systems.

Abstract: Background Short-term exposure to PM2.5 has been associated with respiratory infections, yet evidence on the health effects of its chemical components remains limited. Moreover, as PM2.5 components coexist as complex mixtures, it is unclear whether differences in compositional profiles contribute additional influenza-like illness (ILI) risk beyond the independent effects of individual components. Methods We analyzed weekly ILI surveillance data from 111 sentinel hospitals in 17 cities of Hubei, China (2021-2024), together with weekly PM2.5 concentrations and chemical components (sulfate, nitrate, ammonium, organic matter, and black carbon) from the TAP dataset and meteorological variables from ERA5-Land. To evaluate the associations of individual PM2.5 components and overall compositional profiles with ILI, we applied K-means clustering to identify distinct PM2.5 profiles and fitted city-specific quasi-Poisson models, which were then pooled using DerSimonian-Laird method. Results Among 2,804,416 ILI cases, PM2.5 mass was positively associated with weekly ILI, with a pooled RR of 1.041 (95% CI: 1.028-1.054) per 10 µg/m3 increase. Positive associations were also observed for sulfate, nitrate, ammonium, OM, and BC. Three compositional profiles were identified (SID: secondary inorganic-dominated, MOD: mixed organic-dominated; ORM: Other-rich mixed). After adjustment for PM2.5 and its component concentrations, exposure to ORM was associated with an extra 19.6% higher ILI risk than SID (RR = 1.196, 95% CI: 1.096-1.305), while exposure to the MOD was associated with an 8.9% higher ILI risk (RR = 1.089, 95% CI: 1.008-1.176). Conclusions These findings suggest that PM2.5 related ILI risk may depends not only on overall mass concentration and individual components, but also on compositional profiles. Incorporating PM2.5 mixture heterogeneity may improve assessment of air pollution related respiratory health risks.

Abstract: Modular decision systems expose multiple operating points, but downstream utility can vary by regime. Portfolio construction is a useful setting because routing, aggregation, and allocation can help or hurt depending on market structure. We instantiate a hierarchy with three operating points: direct optimizer, routed consensus, and alpha-augmented optimizer. Across universes, these modes are not uniformly ranked. Routing helps when dispersion/decorrelation is high; direct optimization is safer in low-signal settings; alpha augmentation helps in concentrated signal-rich settings. We identify the market characteristics (cross-sectional dispersion, concentration, forecast signal density) that predict which mode dominates. A rolling adaptive meta-policy that selects among operating points based on recent performance achieves competitive or superior risk-return profiles without foreknowledge of the optimal mode. We validate against classical baselines demonstrate that the operating-point structure persists across time frequencies from 15-minute bars to daily rebalancing, and confirm robustness under realistic transaction costs (0--15 bps). More broadly, our results suggest for hierarchical decision systems: the key is not to find one universally best configuration, but to characterize when each operating mode is most effective. To support future research and ensure reproducibility, we make source code publicly available at https://github.com/kouzhizhuo/Regime-Adaptive-Portfolio-Agents.

Abstract: The effect of atom size on the shock wave structure in a binary monatomic gas mix with Rydberg atoms has been investigated. The problem was solved numerically using the system of hydrodynamic equations in Argon gas, for the atom size ratios between 2 and 100, T = 1500 K, and the density between 10¹⁷ and 10²⁰ m^-3. It was found that the presence of larger size atoms in the mix results in the shock front splitting that is on the order of mean free path for this component. The results can be of interest in supersonic plasma dynamics and in astrophysics studying shock waves in the environments where high-n Rydberg states are present.

Abstract: In this paper, we propose two novel defective survival models within the Gamma–G family: the Defective Gamma–Gompertz and the Defective Gamma–Da- gum distributions. Unlike classical mixture cure models, our formulation incorporates the cure fraction directly into the survival function through the defective property of the baseline distribution, avoiding the need for an explicit mixing parameter. The motivation for these new models lies in the limited set of defective distributions currently available, despite the increasing demand for flexible and parsimonious cure rate models in biomedical applications. By extending the defective property to the Gamma–G construction, our approach fills this methodological gap while providing models that are both interpretable and computationally efficient. We show that the Gamma–G construction preserves defectiveness whenever the baseline distribution is defective, thus establishing a coherent theoretical foundation. Both models allow covariate effects through regression structures on shape, scale, and, in the case of the Gamma–Dagum distribution, on the cure-fraction parameter, resulting in parsimonious and interpretable specifications. Parameters are estimated via maximum likelihood, and an extensive Monte Carlo study confirms estimator consistency and accurate coverage in finite samples. The practical relevance of the models is illustrated with two large clinical datasets on melanoma and cervical cancer from the São Paulo Cancer Registry. Results reveal that the proposed models not only provide superior goodness-of-fit but also offer clearer insights into long-term survival compared to traditional cure-rate approaches. Overal, this work introduces a unifying and flexible framework for defective survival models, extending their applicability and delivering practical improvements over existing cure models.

Abstract: We introduce a new method of dimensional analysis based on complete systems of units, such as the metric and Planck systems, in which fundamental dimensionless constants arise naturally. In fact, it is the reformulated Planck system that communicates its dimensionless constants to the metric or any other system. The method reveals additional complex dynamical scales and physical effects beyond those amenable to conventional dimensional analysis. We formulate our strategy in simple settings involving pairs of seemingly unrelated constants, and then we extend the analysis to more complicated cases involving combinations of three to five well-known universal constants. In constructions involving several unrelated constants, the method captures increasingly complex effects and places two or more disparate physics areas into a single framework connecting them by never-before-seen combinations of fundamental dimensionless constants, such as the fine-structure constant and the gravitational coupling constant. Thus, this method provides an alternative pathway to unified descriptions of fundamental interactions that have so far eluded a consistent theoretical formulation.

Abstract: Background: Plasma biomarkers are widely promoted as scalable tools for staging Alzheimer’s disease (AD), yet head-to-head comparisons against the clinical scales used to define diagnostic labels remain scarce. Reported gains from machine-learning fusion of clinical and biomarker features may therefore reflect label circularity rather than biological signal. Methods: From the Alzheimer’s Disease Neuroimaging Initiative (ADNI), we assembled 655 participants (CN = 296, MCI = 168, AD = 191) with concurrent plasma biomarkers (pT217, Aβ42/40, NfL, GFAP), clinical scales (MMSE, CDR-SB, FAQ), APOE genotype, and demographics. Three pre-specified feature sets (clinical-only, biomarker-plus-demographic-genetic, and full fusion) were compared across four classifiers (Logistic Regression, SVM, Random Forest, XGBoost) using repeated nested cross-validation (5-fold × 3 outer, 5-fold inner) with balanced class weighting. External validation used the Center for Neurodegeneration and Translational Neuroscience (CNTN) cohort (n=130). Results: Clinical scales alone reached a three-class AUC-OvR of 0.9539±0.0041, and fusion reached 0.9559±0.0046, an indistinguishable gain. Because MMSE, CDR-SB, and FAQ partly determine ADNI diagnostic labels, both estimates are circularity-inflated upper bounds. Independently of this circularity, the plasma-plus-demographic-genetic model still achieved AUC-OvR =0.7455±0.0150, with pT217 the dominant contributor. Pairwise discrimination was excellent for CN vs. AD (1.0000) and MCI vs. AD (0.9739), but markedly weaker for CN vs. MCI (0.9302 fused, 0.69 plasma-only). The reduced biomarker model transferred to CNTN with AUC-OvR =0.702 (95% CI 0.635–0.764). Conclusions: Apparent fusion gains in ADNI are largely a consequence of label circularity. After removing the circular clinical features, plasma pT217 supports three-class CN/MCI/AD screening at AUC ≈0.74 internally and 0.70 externally, which establishes a realistic performance ceiling for blood-based AD staging. MCI detection remains the principal bottleneck.

Abstract:

Background: Vitamin D–binding protein (DBP) is the principal carrier of circulating 25-hydroxyvitamin D3 [25(OH)D₃] and is independently synthesized by the neonate. Whether neonatal DBP at birth adds predictive value beyond baseline 25(OH)D₃ for supplementation response remains unclear. Methods: This single-center prospective cohort study enrolled 101 neonates. Neonates with 25(OH)D₃ <20 ng/mL (supplementation-response cohort; n = 59: 29 preterm, 30 term) received 800 IU/day oral cholecalciferol for 8 weeks; neonates with 25(OH)D₃ ≥20 ng/mL served as baseline reference controls (n = 42). Serum 25(OH)D₃ and DBP were measured at baseline and week 8 in the supplementation-response cohort. Results: Median baseline 25(OH)D₃ was 8.60 [6.70–12.05] ng/mL and median baseline DBP was 4.97 [3.70–8.19] µg/mL. After supplementation, 25(OH)D₃ increased significantly (median Δ = 17.70 ng/mL; p <0.001), with 55/59 (93.2%) achieving sufficiency. In multivariable regression, gestational age was the strongest independent predictor of Δ25(OH)D3 (β = −0.440, p = 0.001), followed by baseline 25(OH)D3 (β = −0.314, p = 0.015); baseline DBP was not significant (β = 0.072, p = 0.551). Conclusions: Baseline DBP did not independently predict supplementation response. Lower gestational age and lower baseline 25(OH)D₃ were associated with greater increases in 25(OH)D₃ after supplementation, whereas baseline DBP provided no additional predictive value. Supplementation with 800 IU/day for 8 weeks was effective across gestational-age categories. Routine DBP measurement does not appear to provide additional clinical value for guiding neonatal vitamin D supplementation.

Abstract: The paper examines stationary and dynamic gametheoretic models of cooperation formation among agents based on mutual trust. The dynamic model confirms that trust is a critically important factor in economic interactions. Mutual trust can significantly increase the payoffs for all agents, but its achievement is possible only for a certain class of input parameters of the model and requires a significant return on each agent’s investment in the public good. The model uses the value of social trust to describe social norms accepted in society. The study of the dynamic model was conducted based on Nash equilibrium in the static case and via simulation modelling for heterogeneous agents with different objective functionals. In the model, individual states act as agents. It is shown that the payoff of states directly depends on the degree of trust between them and can grow significantly in the presence of trust. Simulation experiments were conducted in the dynamic case, and the results were analysed.

Abstract: Background/Objectives: Chromosomal microarray analysis (CMA) is an essential tool in modern cytogenetics for detecting copy number alterations and copy-neutral loss of heterozygosity (CN-LOH). As optical genome mapping (OGM) emerges as a potential replacement for traditional cytogenetic methods, the extent to which CMA remains necessary in routine diagnostic workflows remains to be elucidated. Methods: We retrospectively reviewed 53 primary neoplastic cases in which CMA identified one or more CN-LOH events. Event size, genomic content, and correlation with next-generation sequencing (NGS) findings were assessed. A separate cohort of newly diagnosed B-cell acute lymphoblastic leukemia (B-ALL) was analyzed to evaluate disease-specific CN-LOH frequency. Results: Nearly half of CN-LOH events detected were &lt;25 Mb, below the current detection threshold of OGM. Many encompassed clinically relevant genes, including FLT3, JAK2, TET2, TP53, and RUNX1. Additionally, two-thirds of cases harbored pathogenic or likely pathogenic variants by NGS within the corresponding CN-LOH regions, further underscoring the clinical value of detecting these copy-neutral events. In contrast, CN-LOH was uncommon in B-ALL, and most alterations identified by CMA would be detectable by OGM. Many of these patients also harbored complex structural rearrangements that required multiple conventional assays for full characterization; these could be resolved by OGM in a single analysis. Conclusions: Our findings indicate that although OGM excels at resolving complex structural variants, CMA remains essential for detecting copy-neutral events. Until OGM achieves improved sensitivity for CN-LOH, an integrated approach utilizing conventional cytogenetics, CMA, NGS, and OGM provides the most reliable framework for comprehensive genomic assessment across cancer types.

Abstract: Marginal ice zone (MIZ) is a transitional region between dense pack ice and open water. It is a highly dynamic zone under strong interactions between the atmosphere, ocean, sea ice and waves, playing a crucial role in the polar climate and ecosystem. Accurate determination of MIZ is therefore essential for advancing our understanding, modeling and prediction of the polar climate system. In this paper, we introduce and apply a suite of MIZ-related metrics to evaluate the performance of four satellite-derived high-resolution operational sea ice concentration (SIC) products in determination of the MIZs around Svalbard, using the Norwegian ice chart as reference. The metrics used for evaluation include sea ice extent (SIE), MIZ extent (MIZE), length of ice edge (LIE), integrated ice edge error (IIEE), integrated MIZ error (IME), ice edge distance error (IEDE), and MIZ width error (MWE). The evaluation is based on three years of daily SIC data (2023-2025) from four products, including the Bremen AMSR2 SIC data from the University of Bremen (Bremen SIC), the Resolution-enhanced AMSR2 SIC (RE SIC) and Multisensor SIC products (Multisensor SIC) from the Norwegian Meteorological Institute, and the Automated Sea Ice Product (ASIP) from the Copernicus Marine Environment and Monitoring Service (CMEMS) (ASIP SIC). To be consistent with the Norwegian ice chart, the MIZ is defined as MIZ70 and MIZ90, corresponding to SIC thresholds of up to 0.7 and 0.9, respectively. IEDE and MWE are calculated using two types of LIE, the reference LIE (LIEr) from the Norwegian ice chart and the average LIE (LIEa) by averaging the ice chart LIE and the concerned LIE from the four satellite products. The results demonstrate that all four satellite SIC products generally capture the evolution of the sea ice conditions around Svalbard well, but differ in their accuracy when determining the ice edge and MIZs. The Bremen SIC product tends to overlook areas with low SIC, leading to a significant underestimation of SIE and a large IIEE. However, it provides an overall close agreement with the ice chart for the MIZ90 metrics (MIZE, IME and MWE). The RE and ASIP SIC products exhibit strong performance in capturing the ice edge and total SIE, with the ASIP product particularly excelling in accurately representing the ice edge and MIZ70. The Multisensor provides the closest agreement with the ice chart for the IME90, MWE90 and MWEa70, and ranks as the second-best product for the IIEE, IME70 and MWEr70. These results suggest that SAR and Low-frequency AMSR2 channels are especially effective for capturing the lower bounds of the MIZ, while high-frequency channels are more suitable for defining the upper bounds. Despite these strengths, the complex summer surface conditions pose significant challenges for satellite sensors in determining the ice edge and the MIZ, resulting in higher IEDE and MWE values during this period. These results highlight both the capabilities and limitations of satellite-based data in determining the MIZ, particularly under challenging summer conditions. Accurate determination of MIZs may require significant advancements in satellite observation technologies, retrieval algorithms, and more robust methods for integrating multiple sources.

Abstract: The sealing behavior of fracture-filling minerals in the near-field of the deep geological repository (DGR) is critical for the safe disposal of high-level radioactive waste (HLW). In granite host rocks, natural fractures are often filled with polymineralic assemblages of calcite, quartz, and clay minerals; however, their coupled compaction–pressure solution mechanisms under thermal–hydraulic–mechanical–chemical (THMC) conditions remain poorly understood. In this study, 12 fracture sealing experiments were conducted on Beishan granite and its typical fracture fillings at 90 °C and 15 MPa effective stress, using different pore fluids and systematically varying grain size (75–250 μm), mineral proportions, and clay content. The results demonstrate that pressure solution–precipitation of calcite in saturated CaCO3 solution is the key mechanism for long-term fracture sealing, achieving a compaction strain of 24.6%—substantially higher than those obtained in deionized water (20.6%) and under dry conditions (14.8%). Fine-grained calcite compacts more effectively than its coarse-grained counterpart, reaching a porosity as low as 4.8%; rigid quartz accelerates calcite pressure solution via stress concentration at grain contacts; and a moderate amount of clay minerals (~20 wt%) further reduces porosity to 2.1% through lubrication and micropore filling. The study reveals a multi-stage process transitioning from mechanical compaction to pressure solution–precipitation, and a synergistic sealing mechanism dominated by calcite compaction–pressure solution, augmented by quartz stress transfer and clay lubrication. These findings revise the traditional monomineralic understanding and provide a scientific basis for safety assessment of HLW disposal and the design of natural sealing materials.

Abstract: Introduction: Transepithelial photorefractive keratectomy (Trans-PRK) offers superior re-epithelialization and visual recovery. This study evaluates the impact of preoperative refractive status on clinical outcomes and identifies prognostic factors across varying myopic severities. Methods: This retrospective observational study included 125 eyes [64 patients; age > 20 years; best-corrected visual acuity (BCVA) ≥ 20/25] that underwent Trans-PRK between March and December 2022. Patients were stratified into low myopia (LM: > -5.0 D), moderate-to-high myopia (MHM: ≤ -5.0 D to > -8.0 D) extremely high myopia (EHM: ≤ -8.0 D) groups. Analysis focused on preoperative refraction, intraoperative parameters, postoperative uncorrected visual acuity (UCVA) and corneal conditions of superficial punctate keratitis (SPKs) and haze. Results: The mean age was 30.20 ± 6.34 years, with a mean initial manifest sphere (MS) of -6.42 ± 2.27 Diopter (D) overall and -3.73 ± 0.15 D, -6.28 ± 0.13 D and -9.17 ± 0.15 D in the LM, MHM and EHM groups, respectively. At a mean follow-up of 6.69 ± 3.73 months, the overall mean final manifest spherical equivalent (MSE) was -0.12 ± 0.73 D and mean final UCVA was 0.01 [Snellen equivalent (SE), 205/200] ± 0.08 logMAR. Predictability was 94.4%, 88.88% and 94.3% for the final MS ≤ -1.0 D, final MSE ≤ -1.0 D and UCVA ≥ 0.8, respectively. In the LM and MHM groups, cycloplegic and subjective refractions showed the highest concordance with emmetropia, whereas initial manifest refractions were most accurate for the EHM group. Corneal SPK incidence declined from 32.2% (1 month) to 1.6% (6 months), primarily localized to EHM eyes. Corneal haze peaked at 28.2% at three months before receding to 9.4% by 6 months. Conclusions: Refractive and visual stability were achieved by the third month for the LM and MHM groups, whereas the EHM group required six months to reach both refractive and visual plateaus. Despite transiently higher rates of corneal SPKs and haze in EHM eyes, final visual outcomes remained excellent (UCVA 18/20).

Abstract: Active asset managers are increasingly including cryptocurrencies in their alternative asset allocations, highlighting their speculative and volatile nature. The aim of this research is to examine trends in the returns and volatility of cryptocurrencies while accounting for the depegging of stablecoins driven by speculative trading macroeconomic shocks, and technological shifts. It builds a sample, by market capitalisation, using data from the daily closing prices of Bitcoin (BTC), Ethereum (ETH), Binance (BNB), and Ripple (XRP), two fiat-backed stablecoins (USDT and USDC) and a cryptocurrency-collateralised stablecoin (DAI). As a first step, Granger causality tests were applied to examine the influence of stablecoin depegging events on crypto returns during financial market stress. The results indicate that DAI exhibits the most consistent Granger-causal relationship with cryptocurrency returns; whereas, the predictive power of USDT and USDC depegging events varies across assets. The analysis was extended by modelling volatility using an EGARCH-X model to study whether depegs also affect crypto during periods of market stress. In this case, the evidence for statistically significant effects is limited. Nevertheless, in the instances where significance is detected, the results are consistently linked to USDC.

Abstract: To identify the key factors influencing the cracking behavior of fully ceramic microencapsulated (FCM) fuel, this study employed the MOOSE multi-physics coupling platform and the cohesive phase-field fracture theory to simulate crack initiation and propagation in FCM fuel, with particular attention to the effects of particle spacing and residual pore in the matrix. Results showed that during early irradiation stages, in the absence of matrix defects, particle spacing had minimal influence on the distribution of the maximum principal stress. However, when residual pore was present in the SiC matrix, significant stress concentration occurred at the porosity sites, where the maximum principal stress was localized. Smaller particle spacing promoted crack initiation in the SiC matrix between adjacent particles and led to a higher number of cracks under the same fast neutron fluence. In the presence of residual pore, crack nucleation occurred at porosity sites even at low neutron fluence; at a fluence of 2.3 dpa, through-thickness cracks formed in FCM fuel containing residual pore, resulting in the loss of fission product containment capability.

Abstract: Federated learning lets organizations train a shared model without pooling private data. The standard method, Federated Averaging, requires all participants to use the same input features, a condition that fails in cross-sector phishing detection, where banks analyze URL structure and hospitals analyze email content. We present RankBridge, a system that groups participants by comparing ranked lists of SHapley Additive exPlanations (SHAP) feature importance rather than model weights or gradients. Each participant trains a local LightGBM model, extracts the top-K features by SHAP importance, and sends only 60 bytes of ranked indices to a central server. The server applies rank correlation and Ward’s hierarchical clustering to identify similarly threatened organizations, then combines models only within each discovered group. Across 32 participants in five organization types, RankBridge achieves F1 =0.854 on synthetic data and F1 =0.775 on real phishing data. Federated Averaging collapses to F1 =0.278 on the same data. RankBridge recovers the correct organizational groupings with Normalized Mutual Information (NMI) =0.978 while each participant transmits roughly 10,000× less data per round than a full model upload.