Abstract: ANN-to-SNN conversion is an important approach for obtaining high-performance spiking neural networks (SNNs), yet current conversion methodologies predominantly focus on dense architectures, overlooking the structural sparsity that characterizes biological neural systems. In this work, we propose a sparse ANN-to-SNN conversion framework that inherits learned sparse topology into SNNs. We instantiate this framework with Cannistraci-Hebb Training (CHT), a brain-inspired dynamic sparse training method. Across CNN experiments on CIFAR-10, CIFAR-100, and DVSGesture, and ViT experiments on ImageNet-1K, SNNs produced by our framework achieve near-dense performance while outperforming SNNs obtained from static pruning and sparse-training baselines in accuracy. Compared with dense SNN counterparts, our framework also reduces energy consumption in all evaluated settings. Beyond these accuracy-efficiency gains, the resulting sparse SNNs are characterized by biologically plausible structural properties, including meta-depth and scale-free organization. These results suggest that learned sparse topology is a useful design dimension for ANN-to-SNN conversion, enabling high-accuracy sparse SNNs with substantially reduced energy consumption.

Abstract: China’s pollutant discharge permit system mandates total-quantity emission control for industrial volatile organic compounds (VOCs), yet the actual utilization of permitted capacity remains poorly studied. This study developed an “emission idle rate” (IR = 1 − actual/permitted emissions) framework and applied it to 130 chemical enterprises across three cities in Jiangsu Province using 2020–2024 panel data. The mean idle rate reached 78.1%, with no significant inter-city differences (H = 0.96, p = 0.619), attributable to both production underutilization and systematic over-estimation of emission ceilings inherent in the design-capacity-based permit methodology. Ward hierarchical clustering revealed three emission behavioral patterns: Persistent Surplus (n = 74, IR = 0.95), Declining Surplus (n = 32, IR = 0.69), and Growing Surplus (n = 19, IR = 0.59), exhibiting distinct idle rate levels and temporal trajectories. Cluster differentiation was significantly associated only with production-side emission characteristics, while enterprise economic variables showed no significant effects. The estimated tradeable emission surplus reached 668.3 t/a, though its realization faces transaction cost barriers including the lack of standardized transfer mechanisms and formal VOC trading infrastructure. A quadrant-based strategy matrix integrating idle rate levels with temporal trends is proposed for differentiated permit management.

Abstract: Frequency, as a physical quantity that describes the rate at which periodic events occur, is a crucial perspective and component, and can help observe and recognize the world via versatile frequency transforms. Initially, built on Fourier analysis theory, it played an important role primarily in the field of signal processing, and has gradually become an indispensable part of deep learning to solve complex problems. Deep learning in the frequency domain (a.k.a. Fourier domain), which we called \textbf{F}requency-principled \textbf{D}eep \textbf{L}earning (FDL), has been extensively employed in a wide range of scenarios owing to its compelling advantages, such as global receptive field, high computational efficiency, inherent data decomposition and explainability. Deep neural networks also exhibit certain properties from a frequency-domain perspective, which provides valuable insights for powerful model design and refinement. Despite growing attention to frequency-domain approaches in deep learning and computer vision, the absence of a systematic synthesis makes it difficult to grasp the current landscape, identify the core methodologies, perceive the challenge, and chart a course for future research. Moreover, a comprehensive explanation for why introducing frequency-domain methods contributes to problem-solving is still lacking. This survey aims to provide a comprehensive and structured overview of frequency-principled vision and learning to address this gap. Unlike previous reviews that may focus on isolated aspects, our work seeks to connect and systematize the field through a unified taxonomy. Specifically, we conduct a systematic survey and analysis of existing literature from multiple perspectives: frequency principle (theory), implementations (algorithms), applications, challenges and future frontiers of FDL across various tasks.

Abstract: The Kerimov-Alekberli (KA) framework (Karimov and Alekberli, 2026) detects imminent collapse in complex systems by monitoring KL-divergence accumulation relative to a stable reference distribution via a first-passage time (FPT) trigger. Prior work established that incorporating directional asymmetry accelerates detection fourfold. The present paper extends the KA model with three additional structural] components: (1)~directional asymmetry $A(t)$; (2)~temporal memory $M(t)$ via lag-1 autocorrelation deviation; and (3)~a symmetrized entropy production rate proxy $sighat(t)$. A multi-scale detection architecture is introduced, separating wide-window ($\Phi$, $W_\Phi=40$) and narrow-window ($A$, $sighat$, $W_{\mathrm{fast}}=12$) components to ensure mechanistic independence. Monte Carlo validation across three collapse scenarios ($N=200$ each, $FAR=5\%$) yields scenario-dependent gains: $15.9\times$ in Hopf bifurcation, $1.02\times$ in 3-phase drift, $0.98\times$ in TAR model. A 300-sample Dirichlet weight search identifies optimal weights $w^*=[\Phi:0.220,\,A:0.425,\,M:0.269,\,\hat{\sigma}:0.086]$, with default weights achieving $98.7\%$ of optimal. Ablation study confirms $M(t)$ provides the largest marginal gain ($+41.1$ steps in Hopf). Phase-randomized surrogate testing confirms the Hopf memory gain is not a calibration artifact (Mann-Whitney $p<0.0001$, rank-biserial $r=0.952$). Bootstrap $95\%$ CI for composite lead time: $[50.1,\,59.3]$ steps. Cohen's $d=2.65$ (very large) for the memory extension. Empirically-calibrated real-domain validation demonstrates: BTC flash crash composite achieves $24.7$ days lead ($1.08\times$, $100\%$ DR); ICU sepsis onset composite achieves $12.1$ hours lead ($+2.0$,h absolute gain, $82\%$ DR). The central finding is that gain is mechanism-dependent: memory $M(t)$ is most valuable in oscillatory/CSD systems; KL divergence is near-sufficient for distributional-shift collapses. The EPR proxy $sighat(t)$ is grounded as a symmetrized KL rate with $O(\Delta t)$ relative error to true Onsager EPR.

Abstract: Background: Early diagnosis in Bipolar Disorder Type I (BDI) is essential for a better outcome. Endophenotypes are an important subtype of biomarkers but to date very few have been discovered in psychiatry. Some previous studies show anxiety-trait as a possible Bipolar Disorder Type I endophenotype, but there is a lack of replication in another populations. Objective: Our study evaluate the presence of anxiety-trait and bipolarity risk in Bipolar Disorder Type I first-degree relatives (FDR). Methods: We evaluated 219 participants (119 healthy controls, 68 healthy FDR and 32 affective unipolar FDR) including socio-demographic data, psychiatric history, Spielberg's State and Trait Anxiety Inventory (STAI), General Health Questionnaire, Beck Depression Scale, and Mood Disorder Questionnaire (MDQ) scores as a measure of bipolarity risk. Results: In our sample, affective unipolar BDI's first degree relatives showed highest scores in STAI anxiety-trait compared to healthy controls and BDI's healthy first degree relatives. STAI-Trait only correlated significantly with bipolarity risk (MDQ) in the unipolar affective first degree relatives subgroup. Conclusion: Unless further investigation is needed, anxiety trait could be a possible BDI's endophenotype candidate since it seems to have high heritability and would confer higher lifetime risk to develop unipolar affective disorder and tendency to bipolarity.

Abstract: Addressing the issues of insufficient adaptability and limited energy efficiency optimization capabilities in traditional tunnel lighting control methods under complex traffic conditions, this paper proposes a dynamic dimming strategy for tunnel lighting based on the Proximal Policy Optimization (PPO) algorithm.First, the tunnel lighting system is modeled as a reinforcement learning environment. A state space integrating multi-dimensional information—including traffic flow, vehicle speed, external brightness, and tunnel section location—is constructed, and a continuous action space is designed to enable precise dimming control for each functional section. Based on this, a multi-objective reward function is established that integrates brightness tracking error, energy consumption optimization, control stability, and environmental adaptability to guide the agent in learning the optimal dimming strategy.Subsequently, model training and experimental validation were conducted using actual tunnel operation data.Experimental results indicate that, compared to traditional L20 control strategies, the proposed method achieves smoother brightness regulation and higher zone control accuracy while ensuring driving safety and visual comfort, and demonstrates significant energy-saving advantages during periods of high lighting demand. In summary, the dynamic dimming strategy based on the PPO algorithm shows promising application prospects and engineering value in intelligent tunnel lighting systems.

Abstract: An elementary and self-contained approach to the Euler--Mascheroni constant $\gamma$ is presented, based solely on Simpson's quadrature rule and the convexity of the function $f(x)=1/x$. By introducing Simpson-type weighted harmonic sums, local logarithmic increments are approximated by simple finite linear combinations of reciprocal integers. Sharp two-sided inequalities, derived from monotonicity and convexity, yield explicit control of the quadrature error and provide a purely numerical proof of the classical limit defining $\gamma$, without recourse to the Euler--Maclaurin summation formula.A key structural outcome of this framework is a decomposition $\gamma = ( \log [2] + 1 )/3 + \delta$, where the constant $\delta$ arises naturally as the limit of a rational sequence associated with Simpson-regularized harmonic sums. This formulation isolates a dominant oscillatory component and leads to a sequence with markedly faster convergence.This work highlights an unexpected connection between elementary numerical quadrature and one of the fundamental constants of analysis.

Abstract: To address the complex spatiotemporal dependencies and dynamically evolving spatial relationships in tunnel traffic flow prediction, a macro–micro collaborative two-stage prediction method is proposed.The Grey Wolf Optimizer (GWO) is first employed to optimize the GRU model for predicting incoming traffic flow at the tunnel entrance, providing reliable macro-level input for subsequent modeling.Based on this, a spatiotemporal graph structure is constructed, and an FSE-ST-GCN model integrating an adaptive adjacency matrix with spatial and channel attention mechanisms is developed to capture dynamic spatial dependencies and enhance key feature representation.Experiments are conducted using real-world traffic flow data collected from the Shizuizi Tunnel on the Jilin–Caoshi Expressway. The results show that the proposed method outperforms baseline models in terms of MAE, RMSE, and MAPE, achieving superior prediction accuracy and stability. This work provides effective technical support for refined tunnel traffic management and lighting control.

Abstract: We present a 3D laser-scanning method for fast, accurate dimensional inspection of large high-speed-rail precast box girders. The pipeline uses low-pass filtering plus sequential registration to suppress noise, and voxel filtering with curvature-aware enhancement to reduce point-cloud size by 3–5× while preserving key geometry. Reconstruction employs K-nearest-neighbors and PCA to detect boundaries and curvature jumps, B-spline fitting with moving least squares for surface completion, and CSS corner detection to extract key dimensions at millimeter precision. Field tests report absolute errors ≤2.0 mm versus manual measurement, validating the method for automated, digital acceptance.

Abstract: Nanoscale therapeutics, lipid nanoparticles, extracellular vesicles and other particle-like interfaces are usually designed as material objects, but they rarely meet biological readers as pristine surfaces. After entering blood, lymph, interstitial fluid or inflammatory exudates, they are rapidly rewritten by adsorbed, exchanged and modified biomolecules. Protein-corona research has established that this host-conditioned layer can define biological identity, yet the field still lacks a framework for asking a more demanding question: under what conditions can a body fluid maintain a transport-compatible identity for an incoming interface, and when does that assignment shift toward clearance, complement activation, coagulation, inflammation or delivery failure? This Hypothesis proposes finite-capacity corona governance as a testable framework in which protein-rich body fluids function as biological identity-assignment systems with local, finite and disease-sensitive functional capacity. The framework contains two mechanistic claims and one causal attribution standard. First, under matched particle-core conditions, different physiological or pathological fluids assign reproducible surface-accessible corona identity states. Second, for a defined particle class, fluid state and exposure condition, increasing accessible surface-area burden generates a measurable capacity curve; in vulnerable fluids, specific reader modules may undergo nonlinear switching from shielding or transport-associated identities to opsonin-, complement-, coagulation- or instability-associated identities. Because corona composition alone cannot establish causality, the framework further imposes a gatekeeping standard: a fate output should be called corona-governed only when the retained particle-bound identity transfers with washed particles into a common reader system, remains sensitive to label depletion or rescue, and depends on the predicted biological reader. The proposed metric, C_cap,p,f,e,m = S_crit,p,f,e,m / V_fluid, is not a universal plasma constant but a particle-fluid-exposure-specific functional threshold defined for a pre-specified reader module. This framework is intentionally falsifiable: it would be narrowed if response curves remain purely linear, if disease-fluid effects disappear after soluble-factor control, or if fate is better explained by aggregation, material toxicity, payload pharmacology or reader-independent mechanisms. If supported, finite-capacity corona governance would shift nano-bio interface research from asking only what corona forms to asking how host fluids assign biological identity, how much interface burden that assignment can tolerate, when disease states lower this functional reserve and what causal evidence is required before an altered fate can be attributed to the particle-bound corona.

Abstract:

In many scientific domains, physics-based simulators—programs that compute system behaviour from governing equations, such as density functional theory for materials or fluid dynamics solvers—encode causal mechanisms and can predict system behaviour under hypothetical interventions. Machine learning extracts patterns from observational time series at scale, but those patterns reflect statistical associations ($P(Y \mid X)$), not causal effects ($P(Y \mid \mathrm{do}(X))$): in the presence of latent confounders, the structural VAR is provably non-identifiable from observational data alone (Fact 3.3), and no amount of statistical sophistication can substitute for genuine interventional data. Bridging these two traditions has so far been limited to using simulators for prediction; no existing framework uses them for causal structure discovery in time series. We propose SVAR-FM (Structural VAR with Flow Matching), a framework that treats a physical simulator as a mechanical realization of Pearl's $\mathrm{do}(\cdot)$ operator. Clamping a variable in the simulator physically severs confounding paths, producing interventional data by construction rather than by statistical argument. Conditional Flow Matching then parameterizes the interventional conditionals, enabling nonlinear mechanism learning. This yields four results. (1) The full structural VAR—contemporaneous and lagged edges jointly—becomes identifiable under a coverage condition on the simulator-clampable variables, verifiable a priori from domain knowledge alone (Theorem 4.1). The argument is intrinsic to the time series setting and has no i.i.d.\ counterpart. (2) An end-to-end error bound $|\hat{e}_{i\to j} - e^{*}_{i\to j}| \le O(M^{-1/2}) + O(\delta_{\mathcal{S}}) + O(\varepsilon_{\mathrm{FM}})$ (Theorem 5.2) cleanly separates Monte Carlo sampling, simulator fidelity $\delta_{\mathcal{S}}$, and Flow Matching approximation. A sharp consequence is a sign-flip regime (Corollary 5.5): when $\delta_{\mathcal{S}}$ exceeds a threshold set by the signal magnitude, the estimated causal effect reverses sign—a prediction that the prevailing forward-prediction view of simulators cannot produce. (3) The CausalSim benchmark confirms that SVAR-FM recovers the correct causal sign across four scientific domains (macroeconomics, iabetes, cosmic ray physics, and battery degradation) where observational methods produce sign-reversed estimates due to confounding. (4) A case study in ultrafast laser physics tests the sign-flip prediction by physically varying $\delta_{\mathcal{S}}$ through the accuracy level of a first-principles quantum solver: the low-accuracy setting produces a sign-reversed estimate, while the high-accuracy setting recovers the correct positive slope ($R^2 = 0.983$, zero bias relative to ground truth), providing the first experimental demonstration of a simulator-fidelity-dominated failure mode in causal discovery.

Abstract: Background: Endotracheal intubation is a painful and stressful procedure for neonates, often triggering adverse physiological responses. In 2010, the American Academy of Pediatrics (AAP) recommended premedication for elective intubation in neonates to reduce pain and stress and to increase the chances of success. In recent years, several drugs have emerged as a treatment option to improve comfort and safety of neonates undergoing endotracheal intubation. Worldwide, a range of agents are used. As remifentanil is a short-acting analgesic with sedative effects, it may be a suitable option. Therefore, we evaluated the available evidence regarding benefits and side effects for this drug. Methods: In this narrative review, we describe the benefits and risks of remifentanil as drug of choice for neonatal intubation. Results: Literature search demonstrates a relatively limited number of randomized trials, indicating that current practice is informed by a combination of small clinical trials, observational studies, and pharmacological research. Owing to its unique pharmacokinetic profile, remifentanil is an effective agent in the neonatal population allowing the provision of intense analgesia with a rapid recovery profile and has a good clinical applicability in situations where early extubation of patients is desired following the end of the opioid infusion. Reported side effects include respiratory depression, apnea, bradycardia, and chest wall rigidity, similar to other opioids. Chest wall rigidity appears to be strongly influenced by dosing strategies and the speed of intravenous administration. Conclusions: Remifentanil's unique properties make it a promising option for neonatal intubation. However vigilance and close monitoring is required. Further research is warranted to compare remifentanil with other opioids that have a near similar pharmacological profile (e.g., fentanyl analogues).

Abstract: An anti-contamination agent (Zn/Al–ATMP–LDH) has been synthesized by intercalation and used to correct the abnormal thickening and related operational risks caused by contact contamination between drilling fluids and cement slurries during high-temperature/high-pressure cementing. Experimental results have shown that the agent is chemically stable and exhibits good compatibility with conventional spacer-fluid additives. When compared with the direct addition of amino tris(methylenephosphonic acid) (ATMP), confining ATMP within a layered double hydroxide (LDH) markedly mitigates the retarding effect. At a dosage exceeding 0.3 wt%, the compressive strength of cement stone increases from 0 to 32.84 MPa following curing at 90 °C for 1 day, and continues to develop steadily after 7 days. Following conditioning at 187 °C, 145 MPa and 120 min, the spacer system formulated using the proposed agent as the core component serves to enhance the rheology of the mixed slurry via synergistic adsorption-regulation-dispersion stabilization-controlled release. The mixed slurry maintains stable rheological properties before and after aging with no uncontrolled thickening. When mixing the cement slurry and drilling fluid at a 7:3 volume ratio, the slurry consistency exceeds 60 Bc within 1 h, failing to meet operational requirements. In contrast, the mixed slurry containing the anti-contamination spacer (cement slurry:drilling fluid:spacer = 7:2:1) exhibits a thickening time greater than 300 min, and has been successfully applied in field cementing operations in a well in the Gaomo area.

Abstract: Protein Phosphatase 5 (PP5) is an evolutionarily conserved serine/threonine phosphatase with a unique tetratricopeptide domain. It has been implicated in a wide range of cellular processes in mammals, but its function in plants is unknown. Here we uncovered that Arabidopsis PP5 is required for immunity mediated by the nucleotide-binding leucine-rich repeat immune receptor protein SUMM2. Loss-of-function mutations in PP5 suppress the autoimmune phenotypes caused by the activation of SUMM2 due to the disruption of the MEKK1-MKK1/MKK2-MPK4 kinase cascade. Further biochemical analysis revealed that SUMM2 interacts with Heat Shock Protein 90 and PP5, and SUMM2 level is reduced in pp5 knockout mutant plants, suggesting that PP5 functions as a co-chaperone to regulate the accumulation of SUMM2.

Abstract: Although athlete monitoring can quantify training exposure and athlete status with increasing detail, conversion into daily training decisions remains inconsistent. This structured narrative review synthesizes evidence on training load, neuromuscular readiness, recovery, fatigue interpretation, measurement reliability, applied decision-making, and proposes the LOAD-R framework: a systems model linking Load, Organism response, Adaptive state, Decision, and Re-evaluation. A transparent non-PRISMA strategy was used because the aim was conceptual integration and framework development rather than effect-size pooling. Evidence was organized around field-applicable monitoring domains and their decision value. LOAD-R extends existing approaches by moving beyond single indicators, fixed thresholds, and dashboard alerts. It classifies athlete state into adaptive, functional-overload, underloaded, uncertain, or maladaptive zones, each linked to progress, maintain, modify, deload, or recover decisions. The framework also provides implementation levels and testable predictions. By shifting monitoring from passive data collection toward adaptive decision support, LOAD-R offers a scalable model that may improve decision consistency, reduce maladaptive training responses, and enhance the practical value of athlete monitoring in applied sport settings.

Abstract: Background: Influenza vaccination uptake among healthcare professionals remains subop-timal despite their key role in influencing public vaccination behavior. This study investi-gated motivational and behavioral determinants of influenza vaccination uptake and ad-vocacy among primary healthcare professionals in Greece. Methods: A cross-sectional study was conducted among 304 physicians and pharmacists using an anonymous online questionnaire. Vaccination uptake (2023–2024 season and annual) and motivational and advocacy constructs were assessed using the validated MoVac-flu and MovAd scales. Factor structure was evaluated using confirmatory and ex-ploratory factor analyses. Multivariable logistic regression models were applied to identify predictors of vaccination uptake. Results: The study sample consisted of 304 healthcare professionals of whom 61.2% were physicians and 38.8% were pharmacists. Most of the participants were female (52.6%) and aged 41-60 years (57.6%). Influenza vaccination uptake was 77.6% for the 2023–2024 sea-son and 75.3% for annual vaccination. A two- and a four-factor structure was found for the MoVac-flu (F1: Vax Self-Care, F2: Vax Awareness) and MoVAd (F1: Vax Communica-tion, F2: Vax Influence, F3: Vax Confidence, F4: Vax Choice) scales, respectively. Both scales demonstrated high internal consistency (Cronbach’s alpha coefficient > 0.87). Mo-tivation toward influenza vaccination and vaccination advocacy were high among the participants. Vaccinated participants had higher motivation and advocacy scores than non-vaccinated ones. In multivariable analyses, higher scores on Vax Self-Care (aOR = 3.22, 95% CI: 2.08-4.96, p < 0.001) and Vax Communication (aOR = 1.64, 95% CI: 1.14-2.34, p = 0.007) subscales, demonstrating enhanced motivation and advocacy, respectively, as well as male sex (aOR = 2.35, 95% CI: 1.14-4.83, p = 0.020) were associated with higher odds of annual vaccination. Higher scores on the Vax Self-Care subscale (aOR = 3.66, 95% CI: 2.33-5.77, p < 0.001) were also found to be associated with higher odds of 2023–2024 vaccination uptake, as well as living with vulnerable individuals (aOR = 2.95, 95% CI: 1.18-7.38, p = 0.020) Conclusions: Influenza vaccination uptake among primary healthcare professionals in Greece was relatively high; however, it was strongly driven by intrinsic motivational fac-tors, particularly the perceived personal and public health benefits of vaccination. Com-munication-related competencies also independently contributed to vaccination behavior, highlighting the link between professional practice and personal uptake. In contrast, con-textual factors, such as cohabitation with vulnerable individuals, appeared to exert a more situational and less consistent influence. These findings suggest that interventions focus-ing on strengthening intrinsic motivation and communication skills may lead to more sustained improvements in both vaccination uptake and advocacy among healthcare professionals.

Abstract: The transition toward a circular economy (CE) in the plastic recycling sector requires integrated management frameworks that align technical performance with organizational governance. This study proposes an exploratory diagnostic framework for formalized recycling SMEs, integrating Latent Dirichlet Allocation (LDA) and Random Forest (RF) algorithms. Given the specialized nature of the sector, a purposive sample of 16 ‘pioneer’ SMEs in Bogotá was analyzed. Data were standardized through a 5-point ordinal scale, and the Spearman rank correlation analysis (ρ≥0.85) revealed high internal consistency and structural synchronization. This high correlation reflects the operational homogeneity of the analyzed vanguard rather than a universal statistical generalization. The findings suggest that for these leading firms, circularity is driven by social impact, collaborative networks, and systemic process reengineering. The proposed framework serves as a methodological blueprint for analytical generalization, providing an adaptable diagnostic tool that can be iteratively refined as the sector matures and data availability increases.

Abstract: Background and Objectives: Emerging clinical data indicate that glucose levels in women with type 1 diabetes (T1D) fluctuate throughout the menstrual cycle. However, data concerning the adolescent population remain sparse. This study aimed to compare continuous glucose monitoring (CGM)-derived glycemic metrics between the early follicular and late luteal phases in adolescent girls with T1D in order to establish a clinical basis for phase-specific treatment adjustments. Materials and Methods: Ninety-nine menstrual cycles from 34 female adolescents with T1D were retrospectively analyzed. Time in range (TIR), time above range (TAR), time below range (TBR), mean glucose, and glucose variability were compared between the early follicular and late luteal phases. Additional analyses were performed to evaluate the impact of different insulin delivery methods and to assess glycemic metrics across subgroups by glycemic control. Results: During the late luteal phase, TIR was significantly lower, while TAR and mean glucose were significantly higher compared to the early follicular phase (p < 0.05). Although the continuous subcutaneous insulin infusion (CSII) group achieved better overall glycemic metrics then the multiple daily injections (MDI) group, phase-related fluctuations in TIR and TAR reached significance precisely within the CSII group. Furthermore, glycemic shifts in TAR, TBR and mean glucose were especially pronounced in adolescents with tight glycemic control (HbA1c < 7.0%). Conclusions: The late luteal phase is associated with significant glycemic deterioration in adolescent girls with T1D, particularly in those with optimal glycemic control. These findings underscore the need for targeted education and menstrual-cycle-tailored insulin dose adjustments.

Abstract: Livestock production systems in East Africa depend heavily on forage resources, yet productivity and quality of available forages vary widely across agroecological zones. This study evaluated the influence of soil type and altitude on biomass production, biomass allocation, and forage quality of three improved tropical forage grasses—Massai (Megathyrsus maximus), Mestizo (a Urochloa hybrid blend), and Talisman (Urochloa hybrid)—across five experimental sites in Tanzania and Kenya. Field trials were established using a randomized complete block design with three replicates per site. Measurements included cumulative dry matter yield, root: shoot ratio, and nutritive yield expressed as metabolizable energy and crude protein per hectare. Root: shoot ratios varied significantly among species, soil types, and altitudes, with higher ratios observed in Mestizo and Talisman, clay-loam soils, and high-altitude sites. Biomass production was highest in sandy-loam soil and mid-altitude environments. Massai consistently produced the highest cumulative dry matter yield across locations. Significant genotype × environment interactions influenced both productivity and nutritive yield. Metabolizable energy and crude protein yields varied considerably among sites, emphasizing the importance of site-specific forage selection to maximize biomass production and nutritional value in East African livestock systems.

Abstract: Sugar sorghum has emerged as a highly promising crop for bioethanol production owing to its high biomass yield potential and its remarkable capacity to accumulate fermentable sugars in the stalks. The aim of this study was to evaluate the productivity, sugar accumulation capacity, and bioethanol potential of 12 sugar sorghum accessions under the conditions of Northern Kazakhstan. Field experiments were carried out over the 2024-2025 growing seasons, during which key agronomic and technological traits were assessed, including green biomass yield, juice yield, total soluble solids (Brix), sugar concentration, and theoretical ethanol yield. The analysis of variance revealed that biomass yield was predominantly driven by weather conditions (p < 0.001), whereas sugar concentration was significantly influenced by genotype (p < 0.05). Several genotypes, namely Volonter, Kapital, Sevilya, Flagman, Chayka, and Sauri, consistently exhibited high sugar productivity and bioethanol potential across years, confirming considerable genetic variability in these traits. Storage of stem juice resulted in sugar losses of up to 30.7%, indicating the necessity for rapid processing of raw biomass after harvest. Under laboratory fermentation conditions, juice from the Sevilya genotype (17.9 Brix) achieved a sugar-to-ethanol conversion efficiency of 74.3% relative to the theoretical yield. Overall, the findings confirm the suitability of sugar sorghum for bioethanol production and highlight its strong potential as an energy crop under the arid agroclimatic conditions of Northern Kazakhstan.