Abstract: The objective of this study was to evaluate the effect of tea tree essential oil (Melaleuca alternifolia) on postharvest rots caused by Botrytis cinerea and on the sensory quality of ‘Albion’ strawberries. The experiment was conducted with fruits from a production area in Lages, Santa Catarina, Brazil. The treatments consisted of a control (without essential oil) and tea tree essential oil at 50 and 250 µL L⁻¹, applied by fumigation. After treatment, the fruits were stored for three days. Rot incidence and severity, respiration and ethylene production, flesh penetration force, soluble solids, titratable acidity, skin color, total phenolic compounds, peroxidase activity, and volatile profile were evaluated. Tea tree oil reduced the development of B. cinerea at both concentrations. It also delayed red color development and reduced total phenolic content. At 250 µL L⁻¹, the oil increased respiration rate and reduced peroxidase activity. Ethylene production, titratable acidity, soluble solids, and flesh penetration force were not affected. The volatile profile was altered, with lower levels of characteristic strawberry volatiles and greater presence of terpene compounds typical of tea tree oil, especially at the higher concentration. Thus, tea tree oil reduced disease development but caused undesirable changes in fruit color and aroma profile.

Abstract: Timelike boundaries provide a natural setting for organizing causal admissibility, reference structure, visible deviation, and local measurement content on the same Lorentzian surface. This manuscript develops a timelike--boundary reading of local measurement using the established language of special relativity and quantum mechanics as the reference framework. The central object is a timelike boundary equipped with a boundary observer field and observer--adapted cuts. Lorentzian geometry supplies the admissible causal domain of a candidate event. The boundary cut supplies the local comparison surface on which selected quantities are read relative to a coarse--grained reference structure. A local record appears when a boundary--relative deviation becomes resolvable on that cut. The analysis deliberately focuses on the shared interface rather than on the full individual dynamics of general relativity or quantum mechanics. The common structure considered here is the cut--level assignment of causal admissibility, reference structure, resolved deviation, and local record formation. The key distinction is between causal admissibility and measurement content. The causal cone constrains which prior data or contextual contributions may be relevant for a candidate event; it does not by itself supply the local record. The reference structure is specified internally by the boundary reading and is accessed in local records only through locally resolved deviations from it. Thus measurement does not expose the reference structure directly, but records a deviation relative to it. The resulting framework identifies the observer--adapted cut on a timelike boundary as a potential interface where Lorentzian causal geometry and quantum--mechanical record language can be organized together without replacing the established content of either theory.

Abstract: Because wood anatomical traits and tree-ring features vary with species and climatic regime, cellular-scale measurements complement ring-width chronologies and help with understanding how forests may respond to future environmental change. We developed anatomical chronologies spanning the 1900-2019 period from multi-century old yellow pines (Pinus jeffreyi Balf. and P. ponderosa P & C Laws.) at four sites surrounding the Tahoe Basin of the Sierra Nevada, at the border between Nevada and California, USA. Measurements of earlywood and latewood traits included lumen area, lumen length, lumen width, wall length, wall-to-lumen length ratio, and conductive area. Climate sensitivity was estimated by bootstrapped response functions with precipitation and temperature (monthly and seasonal) from the Global Historical Climate Network interpolated to the site locations. Moisture emerged as the primary control on anatomical trait expression, as significant coefficients involved precipitation rather than temperature. Earlywood lumen size and conductive capacity were associated with late-winter through spring moisture, while cellular wall characteristics were connected with conditions during the growing season. Overall, our study provided new insights into the potential impacts of climatic changes on woody species of remarkable size and longevity in an area that is prized for its natural beauty and scenic mountain landscapes.

Abstract: When a UAV is subjected to GPS spoofing, the attacker manipulates key navigation parameters, such as azimuth (heading) and Earth-centered, Earth-fixed (ECEF) coordinates (X, Y, and Z). This leads to inaccurate positioning, causing the UAV to deviate from its intended flight path. A dual-frame simulation architecture is developed, combining a high-fidelity reference model with a real-time flight simulation model running on embedded hardware to evaluate anti-spoofing robustness. Aerodynamic sensitivity is quantified using a Monte Carlo analysis, which identifies pitch moment, yaw moment, and lateral/vertical force lookup tables as the dominant contributors to estimator divergence. A modular aerodynamic coefficient pipeline is introduced, separating datum coefficients, dynamic damping derivatives, and perturbation layers for controlled uncertainty injection. A two layer perturbation framework is implemented to distinguish natural aerodynamic mismatch from spoofing induced inconsistencies during simulation. Cross links between the real AHRS and the onboard flight sim model provide mutual consistency checks, significantly reducing the impact of malicious sensor manipulation. The proposed architecture enables dynamic lookup table switching and morphing during simulation, supporting realistic degradation, drift, and spoofing scenarios. Results demonstrate that physics based cross validation between models and sensors enhances spoofing detection while maintaining robustness to aerodynamic uncertainty.

Abstract: (1) Background: The MultiPill-Exercise study was an exercise-based lifestyle intervention program for subjects with more than one lifestyle-related disease. Metabolomics-based approaches are an emerging tool in sports medicine, both in basic science and in clinical contexts. In this explorative, hypothesis-generating study, we aimed at defining metabolite patterns in subjects with multimorbidity and to test the predictive value of metabolite patterns for fitness adaptation. (2) Methods: A targeted tandem mass spectrometry (MS) approach on dried spots of capillary blood was employed, analyzing n = 86 subjects. (3) Results: In response to an acute bout of exhaustive exercise, aromatic amino acids (AA) tyrosine (Tyr) and phenylalanine (Phe), total carnitine (Cx) and short- as well as medium-chain acylcarnitines (SCA and MCA) were upregulated. Furthermore, concentrations of most metabolites were higher in males. Multiple regression defined free carnitine (C0), MCA and MCA_Delta, i.e. changes in concentrations during acute exercise, as important predictors of intervention-associated fitness adaptation after six months, explaining 24.8% and 24.4% of absolute and relative power gains associated with a respiratory quotient (RQ) of 1, a submaximal fitness parameter. (4) Conclusions: Our data suggest utility of metabolomics data in subjects with multimorbidity, both in the context of basic research and clinical applications.

Abstract: Background/Objective: Caffeine is one of the most extensively investigated supple-ments worldwide, with evidence showing improvements in physical performance across ingestion doses commonly used in sports nutrition (2–9 mg·kg⁻¹). However, studies report substantial variability in aerobic performance outcomes following caf-feine intake, indicating that acute consumption may produce meaningful ergogenic effects but can also impair performance, with time-trial variation ranging from ap-proximately –3% to +16%. Since higher doses may increase the risk of adverse side ef-fects without offering clear added benefits, this review examined the effects of low (≤3 mg·kg⁻¹), moderate (4–6 mg·kg⁻¹), and high (>6 mg·kg⁻¹) caffeine doses on time-trial performance. Methods: A systematic review and meta-analysis of randomized, place-bo-controlled clinical trials evaluating the effects of anhydrous caffeine on aerobic time-trial outcomes was conducted. Random-effects models were applied due to nota-ble heterogeneity across studies, and risk of bias was assessed using the Cochrane Risk of Bias tool. Results: Forty-eight studies (716 participants) met the inclusion crite-ria. Both low and moderate caffeine doses significantly reduced time-trial completion time relative to placebo. Low doses produced a standardized mean difference of –0.27 (95% CI: –0.44 to –0.11; p = 0.001), whereas moderate doses resulted in an SMD of –0.52 (95% CI: –0.77 to –0.28; p < 0.0001). Conclusion: This is the first meta-analysis to demonstrate that pre-exercise ingestion of low caffeine doses (1.3–3 mg·kg⁻¹) can en-hance generalized aerobic performance. Notably, the use of moderate caffeine doses (4–6 mg·kg⁻¹) appears to produce a more consistent ergogenic effect.

Abstract: The paper shows how the ideas of Archimedes about integrating “mechanical methods” and formal reasoning can be connected with the modern-day use of three computer programs – Wolfram Alpha, Maple, and Excel – in exploring topics from elementary theory of numbers. Explorations deal with subsequences of integer sequences through step-by-step elimination of every other term obtained on the previous step. This process, resembling the sieve of Eratosthenes, is applied to tetrahedral numbers appearing in the social context of the family therapy triangulation method. It is demonstrated that symbolic computations of Wolfram Alpha enable generalization in the construction of the sieves that is confirmed by Maple and a spreadsheet. The paper addresses one of the aims of the special issue by demonstrating the duality of mathematics and technology in the sense that whereas the latter facilitates new approaches to knowledge acquisition, the former can be used to improve the efficiency of computations by reflecting on the results made possible by those approaches. The activities advocate for the value of integrating ancient ideas, digital tools, and elementary number theory in the education of mathematics teachers. Reflective comments by teacher candidates are included as appropriate.

Abstract: Background: Common Variable Immunodeficiency (CVID), a heterogeneous syndrome characterized by hypogammaglobulinemia and defective humoral immunity, is the most prevalent symptomatic primary immunodeficiency. CVID-8 is a monogenic variant due to bi-allelic mutation in the LRBA gene. Individuals carrying a single mutated LRBA allele are considered phenotypically healthy. However, immune dysregulation may arise in certain heterozygous carriers likely via haploinsufficiency or dominant-negative activity. LRBA critically regulates CTLA-4 recycling, directly linking this deficiency to immune checkpoint biology. Case Presentation: We report a 7-year-old female, born to consanguineous Lebanese parents, with a family history of thrombocytopenia, presented with chronic refractory ITP first diagnosed at age 2. The patient was resistant to multiple sequential therapeutic interventions including immunosuppressive agents and splenectomy. Whole exome sequencing (WES) detected compound heterozygous LRBA variants c.7937T>G (p.Ile2646Ser) and c.7046T>A p.Leu2349*, the former is pathogenic associated with CVID-8. Immunological assessment revealed hypogammaglobulinemia with suppressed IgG1, IgG3, IgA and IgM levels. Her latest hospitalization was marked by abdominal pain, impaired consciousness, acute liver injury, coagulopathy, peripheral leukocytosis, and lung infiltrates on imaging, suggesting autoimmune enteropathy complicated by infection. Conclusion: This report raises important questions regarding the clinical impact of heterozygous LRBA variant. It highlights the diagnostic value of WES in refractory cytopenias and inherited immune deficiencies. Abatacept, a CTLA-4-Ig fusion protein, is a promising targeted therapy for LRBA deficiency cases; yet its unavailability in Lebanon impeded its use, emphasizing the critical gap in access to targeted biologics in the MENA region.

Abstract: Rapid detection and localization of liquid fuel spills is critical for first responders assessing fire and health hazards, yet current methods require ground-based sampling or specialized instrumentation, limiting their practicality for wide-area emergency response. We present a drone-based passive colorimetric sensor system using test strips impregnated with Nile red, similar to colored confetti. Nile red is a solvatochromic dye that undergoes distinct visible color transitions upon exposure to different liquids. The dye is embedded within a polymer matrix that minimizes leaching while providing high optical contrast between dry, water-exposed, and fuel-exposed states. The sensor strips exhibit solvent-specific colorimetric responses within one minute of exposure, readily detectable by standard RGB cameras mounted on unmanned aerial vehicles (UAV) at altitudes up to 50 m. Automated classification was validated at 20 m altitude, enabling remote surveillance of contaminated surfaces without specialized equipment. Color-corrected image analysis using Calibrite ColorChecker calibration ensures reliable interpretation under variable field illumination (625–77,000 lux). Systematic laboratory evaluation of twelve fossil and bio-derived fuels revealed characteristic hue shifts that clearly discriminate ethanol-containing gasoline blends from diesel-range fuels. Field validation confirmed localization and classification of fuel-exposed sensors, achieving F1 scores of 0.94 for gasoline and 0.98 for diesel detection with no false positives in the tested scenarios. This cost-effective and scalable approach provides actionable information on both contamination location and fuel type, crucial for rapid hazard assessment in emergency response scenarios.

Abstract: Metabolic syndrome (MetS) is highly prevalent in people living with HIV (PLHIV), but whether their intestinal microbiota differs from that of HIV negative individuals with MetS remains unclear. We conducted a cross-sectional study including 30 virologically suppressed PLHIV with MetS and 30 HIV-negative individuals with MetS. Fecal microbiota composition was assessed by 16S rRNA gene sequencing, and predicted functional profiles were inferred using PICRUSt2 and MetaCyc. PLHIV with MetS exhibited markedly reduced alpha diversity and a clearly distinct beta diversity profile compared with HIV negative MetS, indicating a remodeled community structure. Differential abundance analysis showed enrichment in PLHIV + MetS of Prevotella, Selenomonas, Odoribacter, Christensenellaceae R-7 group, and uncultured Lachnospiraceae, whereas Subdoligranulum and the Ruminococcus gauvreauii group were relatively more abundant in HIV negative MetS. Functional predictions revealed higher representation in PLHIV + MetS of Gram-negative cell envelope and lipopolysaccharide-related pathways, amino acid degradation, and ppGpp biosynthesis, while HIV negative MetS showed comparatively greater saccharolytic potential. Carbohydrate related pathways correlated positively with adiposity and blood pressure, and Prevotella correlated positively with BMI only in PLHIV + MetS. These findings support MetS in chronic treated HIV as a distinct dysbiotic and metabolically adverse intestinal phenotype and highlight the intestinal microbiota as a potential target for microbiome-oriented interventions in this population.

Abstract:

This review provides a comprehensive overview of the genetic diversity and epidemiological potential of Yersinia pestis in Kazakhstan’s natural plague foci, emphasizing the link between genotypic variation and outbreak capacity. Integrating historical epidemiological records with contemporary microbiological and genomic data (including PCR, VNTR/MLVA, SNP analysis, and whole-genome sequencing), we evaluate core and accessory genome variations. The data reveal substantial regional heterogeneity. High-risk desert foci (Caspian and Aral regions) are dominated by the Medievalis biovar, including atypical genovariants lacking canonical markers. Conversely, high-mountain foci (Sarydzhaz, Talas) harbor Antiqua and Talas biovars primarily linked to enzootic circulation. Notably, the Ili River focus exhibits extreme genomic variability, featuring strains with plesiomorphic traits. Furthermore, the widespread distribution of mobile elements like the cryptic plasmid pCKF suggests significant horizontal transfer contributing to pathogen adaptation. Ultimately, Central Asian plague dynamics are driven by complex evolutionary and ecological interactions. Given climate change and expanding human-wildlife interfaces, continuous genomic and ecological surveillance is essential for the early detection of high-risk Y. pestis genovariants and improving public health preparedness.

Abstract: Rational use of medicines (RUM) is a cornerstone of safe, effective, and affordable health care. In Nepal, irrational medicine use-particularly of antibiotics-remains widespread, driven by self-medication, over-the-counter (OTC) sales without prescription, inappropriate prescribing practices, weak regulatory enforcement, and commercial influences. These practices contribute directly to the growing burden of antimicrobial resistance, increased adverse drug reactions (ADR), and avoidable economic costs for households and the health system. This perspective article synthesizes recent evidence from knowledge, attitude, and practice (KAP) studies, national policy experiences, and global trends from the past five years to examine the magnitude and drivers of irrational medicine use in Nepal. Ensuring RUM is a shared responsibility involving individuals, communities, health professionals, regulators, policymakers, pharmaceutical industries, and the media. Strengthening RUM in Nepal requires coordinated regulatory enforcement, antimicrobial stewardship, and community-level behavior change.

Abstract: This paper presents the design, implementation, and evaluation of Shine Bright Like Silver, a mobile application developed for PIOP's Silversmithing Museum in Ioannina, Greece, to document, simulate, and present the sand casting and filigree silversmithing techniques. The proposed solution integrates semantic knowledge graphs, an interactive crafting simulation, and gamified learning elements to externalize and operationalize the implicit procedural knowledge involved in the craft. A user study was conducted with n=26 museum visitors to assess the system's effectiveness in terms of usability, learning outcomes, and early indicators of craft valorization. The findings suggest that combining semantic representation with interactive simulation, embedded within a museum context, offers a scalable and replicable framework for the digital preservation of heritage crafts. The work contributes to the broader discourse on technology-mediated craft education by illustrating how tacit knowledge can be systematically captured and transmitted through mobile digital environments, and by positioning visitor-facing tools within a broader reenactable preservation infrastructure.

Abstract: The proliferation of digital content platforms has rendered personalized recommendation systems a foundational component of modern information retrieval. Classical Matrix Factorization (MF) methods, while computationally tractable, are fundamentally constrained by the linearity of the inner product operator, which prevents them from capturing the non-linear, higher-order dependencies characteristic of real-world user–item interaction spaces. This paper presents a complete end-to-end system embodying Neural Collaborative Filtering (NCF), wherein a Generalized Matrix Factorization (GMF) module and a Multi-Layer Perceptron (MLP) are fused to model both linear and non-linear latent factors simultaneously. Two fully isomorphic implementations are developed: a pedagogical NumPy-based version featuring hand-derived backpropagation, and an optimized PyTorch version leveraging Apple Silicon MPS acceleration. Empirical evaluation on the MovieLens dataset demonstrates that both implementations converge to equivalent final Binary Cross-Entropy losses (0.2257 and 0.2307, respectively), while the PyTorch variant achieves a 3.39× speedup in total training time (3,295 s versus 972 s over 20 epochs). Peak recommendation quality, measured by Hit Ratio at cutoff 10 (HR@10), reaches 0.615 for the PyTorch implementation. The system is deployed as a production-grade microservices architecture comprising a FastAPI gateway, a dedicated PyTorch inference server, a PostgreSQL persistence layer, and a Netflix-style frontend with TMDB poster integration. A hybrid cold-start module supplements the NCF core for new users and items. The findings validate the feasibility of bridging rigorous algorithmic pedagogy with industry-standard deployment practices.

Abstract: Public sector organizations face growing pressure to modernize service delivery through digitalization while ensuring transparency, interoperability, and citizen trust. Although blockchain technology offers promising capabilities for addressing these challenges, the absence of clear architectural guidelines for public sector contexts limits effective adoption. This study proposes BRA-PS, a Blockchain Reference Architecture for Public Sector Citizen-Centric Applications, developed from a realworld digitalization project in Quebec, Canada. The architecture organizes components into six layers (presentation, business, communication, smart contract, blockchain, and data) with cross-cutting concerns addressing governance, access control, security, and monitoring. A key design principle is the publicprivate workflow separation, which enables inter-organizational collaboration while preserving each organization’s operational autonomy and data confidentiality. We validated the architecture through a case study involving a vehicle registration process between two public agencies, supported by a proof-of-concept implementation using Hyperledger Fabric. An Architecture Tradeoff Analysis Method (ATAM) evaluation, conducted with a panel of five domain experts, identified six architectural risks, including IPFS confidentiality exposure and smart contract inflexibility, six non-risks, six sensitivity points, and six trade-offs across three key quality attributes: autonomy, collaboration, and functional suitability. The results confirm that BRA-PS effectively guides implementation decisions and stakeholder alignment. Practical recommendations derived from the evaluation provide actionable guidance for blockchain adoption in public sector services.

Abstract: A function-based definition of a living individual as a central element and specially distinguished level of the phenomenon of life is proposed. According to this definition a living individual is a causally-closed network of five types of subordinate functions: p-processes (Ps), signals (Ss), feedbacks (Fs), re-writing (R(s)) and decoding (D(s)), directed recursively on the superior function: survival and proliferation (copying of its own identity in a possibly large number of copies possibly similar to the original), being self-dependent and self-sufficient in the realization of this purpose in its environment/ecological niche. This definition separates univocally living phenomena from inanimate processes in the case of the contemporary terrestrial life, origin of life on our planet, spontaneously-originated alien life in the universe, artificial and virtual life. According to this definition, life originated in the moment of formation of the first genetic code, including the “key” to this code coupling autocatalytic cycles of proteins (with random amino-acid sequences and divers catalytic abilities) and RNA (R), supported by and supporting proto-metabolism (P) and embedded in coacervate-like bubbles, into hypercycles of proteins and RNA (D) (“everything first” hypothesis). It clearly distinguishes the living individual from other levels of the organization of biological systems, such as living cells and organs, populations, ecosystems and the whole biosphere. Finally, it clearly decides whether such phenomena of doubtful status as viruses and viroids, plasmids, retrotransposons, chromosomes B, prions, cancer cells, mutualistic symbionts, mitochondria, colonial coelenterates, slime molds and eusocial insect(s) (colonies) are living individuals or not, and to what degree.

Abstract: Virtual currencies have become an increasingly prominent feature of the modern financial system, yet their public perception and adoption remain insufficiently understood. This study examines the perception and use of virtual currencies in Slovakia. Its primary objective is to assess public awareness, attitudes, and patterns of use. Data were collected through a structured questionnaire covering knowledge, perception, and engagement with virtual currencies. The responses were analysed using graphical methods and statistical techniques, including the t-test and Pearson’s chi-square test of independence, which were also employed to test the proposed hypotheses. The findings indicate high levels of awareness among younger respondents, but also a limited understanding of the underlying technologies and broader potential of virtual currencies. The results further suggest that many individuals hold only one virtual currency, reflecting either insufficient information or uncertainty about these assets. On the basis of these findings, the study proposes measures to improve the current situation, including targeted educational initiatives, investment in research and development, wider practical applications, enhanced security and trust, and continuous monitoring of market trends. The study offers evidence-based insights into the Slovak context and may also inform policymakers, educators, and financial institutions seeking to promote informed and responsible engagement with virtual currencies.

Abstract: Smallholder aquaculture communities at Musangezhi and Chisola Dams in Kalumbila District, Zambia, face escalating, poorly characterised water-quality threats from the adjacent Trident copper mine, yet no real-time monitoring infrastructure exists at either site. This paper presents the design, deployment, and empirical evaluation of the Resilient AI-Enhanced IoT (RAEI) framework a seven-node, solar-powered LoRaWAN sensor network coupled with a comparative machine-learning suite comprising Random Forest, XGBoost, Long Short-Term Memory (LSTM), and CNN-LSTM Hybrid models—trained on 3,551 ICP-OES heavy-metal observations covering copper (Cu), cobalt (Co), iron (Fe), and lead (Pb). XGBoost achieved the highest predictive performance across all four metals and all four-evaluation metrics, attaining a mean R2 of 0.515 and a mean MAPE of 35.89%, with lead prediction reaching R2 = 0.673. A TinyML-quantised LSTM on ESP32 microcontrollers ensured on-device anomaly alerting despite the loss of cloud connectivity. A 14-day trial field test achieved a composite resilience score of 7.5/10 (Technical: 7.4; Data: 8.3; Operational: 6.8). Desire to adopt the community was 73.3%, with cooperative membership (OR = 3.12, p < 0.001) and mobile-money use (OR = 2.67, p = 0.004) being the most significant factors. The RAEI framework detected 97.9% of contamination events missed by the prevailing quarterly manual-monitoring regime. These results confirm the RAEI framework as a technically viable, economically justified, and community-compatible solution for mining-proximate aquaculture surveillance across sub-Saharan Africa.

Abstract: Background: The number of elderly patients with hepatocellular carcinoma (HCC) has been increasing worldwide. This study aimed to evaluate the clinical outcomes of stereotactic body radiation therapy (SBRT) for patients aged ≥80 years with solitary HCC. Methods: This retrospective study included 117 patients with solitary HCC treated with SBRT between January 2019 and June 2025 at eight institutions in Japan. Patients were categorized into the very elderly group (≥80 years, n = 41) and the control group (<80 years, n = 76). Overall survival (OS), local tumor progression (LTP), and treatment-related toxicities were evaluated. Propensity score matching (PSM) was performed to minimize baseline imbalances between the two groups. Results: The median age was 84 years in the very elderly group and 74 years in the control group. The 3-year OS rates were 79.0% in the very elderly group and 62.8% in the control group, with no statistically significant difference observed (crude cohort, p = 0.072; PSM cohort, p = 0.114). The 3-year LTP rates were 18.9% in the very elderly group and 15.6% in the control group, without significant intergroup differences. Acute toxicities were numerically more frequent in the very elderly group; however, no grade ≥3 toxicities or treatment-related deaths were observed. Changes in ALBI score did not differ significantly between groups (p = 0.785). Conclusions: SBRT may represent a feasible treatment option with an acceptable safety profile in carefully selected very elderly patients with solitary HCC. It may be a reasonable option for this growing patient population.

Abstract: Background: Finerenone is associated with a lower, yet clinically relevant, risk of hyperkalemia compared with steroidal mineralocorticoid receptor antagonists in diabetic kidney disease (DKD) trials. However, real-world data on hyperkalemia and its associated factors are lacking. Methodology: FINE-TURK is a national, observational cohort of DKD patients who were initiated on finerenone. Eligible adults were included; demographic, clinical, and laboratory data were evaluated. The primary outcome (PO) was hyperkalemia risk signal (potassium ≥ 5.0 mEq/L), and the secondary outcome (SO) was clinically meaningful hyperkalemia (potassium ≥ 5.5 mEq/L). Multivariate logistic regression (LR) was used to define features associated with both PO and SO. LR, random forest (RF), gradient boosting, and CatBoost classifiers were used to define important features associated with the PO. Results: 699 patients were included. 259 (37.1%) reached the PO, and 51 (7.3%) reached the SO. Baseline potassium and estimated glomerular filtration rate (eGFR) were the most important variables associated with both outcomes and were consistently identified as the top features across all models. Thiazide use, presence of diabetic retinopathy, and diabetes duration were also associated with the PO. LR demonstrated the highest recall; random forest achieved the highest precision in performance. Discussion: Real-world data suggest that the risk of clinically meaningful hyperkalemia is similar to that in the clinical trials. In parallel with the safety analysis of clinical trials, baseline potassium and eGFR were consistently the most important factors associated with hyperkalemia risk.