Search Results (956)

Education on natural hazards is a non-structural component of Disaster Risk Reduction (DRR), but the evidence base spans curriculum studies, risk-perception research, educational interventions, geospatial approaches and analyses of education-system continuity. This systematic narrative review synthesizes 27 outcome-bearing studies published between 2013 and the partial year 2026. PRISMA 2020 was used as a reporting framework, while PRISMA-S informed a retrospective audit of the search documentation. A structured design-sensitive appraisal recorded evidence family, comparison or temporal structure, outcome directness, permitted inference and principal limitation. The studies were coded into six mutually exclusive primary axes: reviews and frameworks; curriculum and policy; knowledge and risk perception; educational interventions and active methodologies; GIS and geospatial technologies; and educational continuity and system resilience. The included literature suggests that locally situated problems, maps, simulations and inquiry can support knowledge, risk appraisal and preparedness intentions, although demonstrated effects on sustained performance or actual preparedness behavior remain limited. Cross-cutting gaps include weak longitudinal assessment, sparse attention to teacher professional development, limited treatment of indigenous or local knowledge, and no core study centered on learners with disabilities or special educational needs. The review defines critical territorial risk literacy as the capacity to interpret hazard, exposure, vulnerability, capacity and uncertainty through spatial evidence; evaluate their unequal territorial distribution; and translate that understanding into inclusive, proportionate preparedness and collective action. Full article

Arabic remains underrepresented in many task-specific natural language processing (NLP) resources and benchmarks, particularly for specialized user interface (UI) understanding tasks. In mobile commerce, Arabic UI text may contain persuasive or deceptive design cues known as dark patterns; however, Arabic-language dark pattern detection remains largely unexplored. To the best of our knowledge, this paper presents the first ML-based benchmark for Arabic dark pattern detection. We construct a novel annotated dataset of 223 Arabic UI text strings from nine e-commerce mobile applications operating in Saudi Arabia, labeled across five dark pattern categories and a non-dark-pattern class (Cohen’s kappa κ = 0.89). Using a stratified, leakage-free 70/10/20 split with parent-aware paraphrase augmentation applied only to the training partition, we fine-tune five pretrained transformer models: AraBERTv2, MARBERT, mBERT, BERT-base-uncased, and RoBERTa-base. Our primary evaluation is 5-fold cross-validation on the 223 original, non-augmented instances, separate from the augmented training corpus used for the held-out test comparison. Under this evaluation, MARBERT achieves the strongest performance (mean macro-F1 = 0.4230), numerically ahead of AraBERTv2 (0.2998) by a margin that does not reach statistical significance at five folds (p ≈ 0.064), and ahead of mBERT (0.3234); MARBERT significantly outperforms both English-only baselines, and mBERT is numerically stronger than both, though mBERT was not directly tested against them for significance. This suggests pre-training on dialectal, code-switched Arabic may matter more here than Arabic pre-training alone. Per-class analysis shows every model struggles with several minority categories, indicating Arabic dark pattern detection remains genuinely difficult at current data volumes. The annotated dataset is publicly released to support future low-resource Arabic NLP research. Full article

Nanolayered metallic composites exhibit ultra-high strength but suffer from inadequate ductility. Constructing interfacial layers becomes an effective strategy to enhance their strength and ductility. However, extensive experimental studies have focused on the strengthening and toughening mechanism of some special interfacial layers under loading normal to the interface, a systematic understanding of how interfacial layer characteristics influence the mechanical response remains unclear, especially under loading parallel to the interfaces. Here, molecular dynamics simulations using the LAMMPS code with embedded atom method potentials are performed to investigate the tensile deformation of Cu/Ni nanolayered composites with various interfacial layers made by six different FCC metals, i.e., Ag, Al, Au, Pb, Pd, and Pt, under loading parallel to the interfaces. Our simulation results reveal that the strength of composites is governed by the metallic element of interfacial layers. The composites with Pd and Pt interfacial layers exhibit the highest and lowest strength, respectively, showing a maximum strength difference of 1.09 GPa. The strength variation is attributed to the synergistic interplay of multiple characteristics of interfacial layers, rather than from a single dominant factor. Furthermore, a quantitative mapping relationship between the strength of composites and the characteristic parameters of the interfacial layers was established on the basis of the Voigt model and the dislocation nucleation behavior. Accordingly, the strength can be expressed as a function of three decisive determinants: the strain energy density required for dislocation nucleation in the pure metal corresponding to the interfacial layers, the elastic modulus of that pure metal, and a parameter related to the stress concentration level at the interfaces. A higher value of the former two factors, combined with a lower value of the latter, corresponds to a higher strength. Full article

This paper considers the task of assessing the relevance of a procurement procedure for a supplier under conditions of growing volumes of open procurement data and increasing complexity of electronic publication services. It is shown that, for a supplier, relevance cannot be reduced to a match between a procurement notice and a search query or classifier code. In the European procurement context, procedure assessment requires consideration of the procurement object, lot structure, selection and award criteria, European Single Procurement Document requirements, procedure language, place of performance, participation through a specific legal entity, contract terms, data-protection requirements, service obligations, risk signals, and retrospective context. A formal mathematical representation of relevance assessment is developed, and on this basis a cascade agent model is proposed in which relevance is defined as the result of matching a procurement procedure, a supplier profile, and a set of evaluation rules. The model includes specialized agents for query normalization, procedure search, primary filtering, documentation preparation, fact extraction, constraint identification, procurement-object identification, customer analysis, contract-terms analysis, supplier-profile matching, dossier checking, relevance calculation, result interpretation, and final-card generation. The key intermediate result is an analytical procurement dossier containing structured facts, constraints, risk signals, matching results, and evidential links to sources. The proposed model is implemented as a research prototype and evaluated on an active stream of procurement procedures published through TED. The prototype evaluation shows that the cascade can transform a broad and heterogeneous set of retrieved procedures into a multilevel operational funnel comprising a shortlist, comparison pool, manual-review routes, and a monitoring layer. Within the reported case, these outputs demonstrate execution of the intended cascade sequence and generation of explainable user-specific assessments; they are not presented as evidence of ranking accuracy, superiority over simpler alternatives, or generalizability. Full article

Over the past decade, advances in sequencing technologies and computational pipelines enabled the development of personalized cancer vaccines (PCVs). Current PCV strategies primarily target cancer neoantigens generated by non-synonymous DNA mutations, which can result in altered amino acid sequences capable of eliciting tumor-specific immune responses. More recently, a distinct class of tumor-specific antigens (TSA), termed noncanonical or cryptic antigens, has emerged as an additional source of immunogenic targets. Unlike canonical neoantigens, noncanonical antigens typically cannot be identified by tumor/normal whole-exome sequencing, as they do not arise from classical DNA mutations. Instead, they are often associated with less well recognized and/or aberrant processes in the pathways from DNA to human leukocyte antigen (HLA)-presented peptides. Examples include transposable elements, circular RNA, translation of alternative open reading frames and/or long non-coding RNA, among others. Emerging evidence suggests that noncanonical antigens represent a substantial portion of the tumor-specific immunopeptidome and, similar to canonical neoantigens, are absent during thymic selection and can evade central tolerance and elicit T cell responses. Technological advances have increasingly facilitated the identification of noncanonical antigens. Long-read RNA sequencing reveals noncanonical transcripts by improving transcriptome assembly, while ribosome profiling provides genome-wide maps of actively translated regions, facilitating the discovery of peptides from aberrant translation events. Specialized molecular approaches enable enrichment and sequencing of circular RNAs, and immunopeptidomics using mass spectrometry allows for direct characterization of HLA-presented peptides. Together, these technological advances have led to an increasing interest in prioritizing and targeting noncanonical antigens in the next generation of PCVs. This review provides an overview of the diverse origins of TSAs beyond classical neoantigens and discusses emerging approaches that may enable the integration of these antigens in future clinical trials. Full article

Modern applications rely on diverse database types to store information efficiently, but interacting with these systems requires developers to master multiple specialized query languages. This fragments development, as engineers must write complex custom code for each database just to retrieve routine data. While existing big-data tools offer a unified query language, they force all data requests through heavy processing engines, degrading the fast performance required for everyday tasks. To address this problem, we introduce PolyFuseQL, a smart routing layer acting as a universal SQL translator for five popular databases (PostgreSQL, Redis, Neo4j, MongoDB, and Cassandra). Our objective is to allow developers to communicate transparently with diverse systems using standard SQL, completely avoiding the need to write specific NoSQL logic. Methodologically, PolyFuseQL dynamically evaluates standard SQL queries to determine the optimal execution path. It translates and sends simple requests directly to the native databases to ensure high speed while seamlessly routing complex analytical queries to a powerful Apache Spark engine. We evaluated the system’s performance across varying workloads. Results indicate that PolyFuseQL adds almost no delay to simple queries, preserving sub-millisecond response times. Concurrently, it enables users to execute complex operations, such as eight-table joins, on NoSQL databases that lack native relational support. Although a fixed processing delay during initial data translation slightly limits absolute hardware scaling, the system scales reliably for medium-sized analytical tasks. Ultimately, PolyFuseQL provides a unified, easy-to-use SQL interface to diverse NoSQL databases without sacrificing operational speed. Full article

When a system must detect several related conditions over the same inputs, its designer can train a specialized classifier for each condition or one pooled model for all of them. Published comparisons of these designs contradict one another, and certain dataset properties, class imbalance above all, usually take the blame. We re-examine the comparison on code-smell detection, a domain where the contradiction is well documented, testing three ways of building the pooled baseline across five code-smell datasets, four generic benchmarks, four classical classifier families, and a neural network, with matched cross-validation folds and threshold-independent scoring. Much of the disagreement turns out to be manufactured by the comparison itself. The conventional pooled baseline appends a task indicator to the features, which quietly forces linear models, and neural networks conditioned only through the input, to share one set of feature weights across tasks; tree ensembles and softmax classifiers, which carry per-task structure of their own, are unaffected. Fixed-threshold F1 scoring accounts for the rest. Once the pooled model receives per-task weights, the gap closes on every dataset, and controlled positive-rate sweeps show that class imbalance moderates nothing. For tabular task families of this kind, the choice between the designs is an engineering decision, not an accuracy one; we close with a short protocol for making such comparisons fairly. Full article

Earth Observation (EO) analysis increasingly relies on large and heterogeneous satellite datasets, yet developing EO workflows often requires specialized expertise in data selection, geospatial programming, and cloud-based processing. Recent advances in Large Language Models (LLMs) offer new opportunities for natural-language interaction with EO systems, although challenges related to transparency, reproducibility, and domain-specific reasoning remain. This study presents SeaScope, an explainable AI framework that integrates LLMs, Retrieval-Augmented Generation (RAG), scientific knowledge retrieval, and Google Earth Engine (GEE) to transform natural-language requests into transparent and executable EO workflows. The framework combines knowledge retrieval, code generation, cloud execution, provenance tracking, and interactive visualization within a unified environment. A pilot implementation is demonstrated through maritime and coastal monitoring applications, including oil spill detection, vessel monitoring, water quality assessment, floating debris detection, and air quality analysis. Multiple state-of-the-art LLMs are evaluated under both RAG and non-RAG configurations using representative EO case studies. The results indicate substantial differences among model families and show that retrieval augmentation can significantly improve workflow generation quality and reliability for capable models, while providing more limited benefits for smaller models. The proposed framework demonstrates the potential of explainable AI agents to support transparent, reproducible, and scalable EO analysis. Full article

Motor neurons form a highly specialized network composed of α-, β-, and γ-subtypes that coordinate skeletal muscle activity. Motor neuron diseases (MNDs), including amyotrophic lateral sclerosis (ALS) and spinal muscular atrophy (SMA), are characterized by the progressive degeneration of this network, resulting in motor dysfunction. Emerging evidence underscores the significant roles of long non-coding RNAs (lncRNAs) in motor neuron development and disease. However, only a few have been experimentally confirmed as true ceRNA regulators, highlighting the need to differentiate validated mechanisms from mere associations or predictions. This review summarizes the regulatory roles of lncRNA-associated ceRNA networks in motor neuron development, evaluates the evidence for their involvement in MNDs, and explores their potential impact on disease progression. It also addresses current challenges, knowledge gaps, and future research directions for understanding ceRNA-mediated mechanisms and developing therapeutic strategies for MNDs. Full article

Breast cancer (BC) is a heterogeneous disease influenced by genetic and regulatory mechanisms. Despite this, X-linked non-coding variants remain underexplored, particularly in admixed Latin American populations. Therefore, this study aimed to characterize the non-coding variant rs5945619 at Xp11.22 in Ecuadorian mestizo women with BC and to assess its potential regulatory effects on nearby genes through in silico cis-eQTL analysis. A prospective observational case series was conducted in 21 Ecuadorian mestizo women with histologically confirmed BC. Tumor DNA was extracted and analyzed using the Illumina TruSight Cancer Sequencing Panel. The rs5945619 variant was identified from sequencing data and compared with reference allele frequencies from dbSNP, ALFA, and the 1000 Genomes Project. Regulatory potential was assessed using RegulomeDB v2.2, and tissue-specific cis-eQTL associations were explored using GTEx. All 21 patients carried the genetic variant rs5945619 in a heterozygous state. The mean age at diagnosis was 54.5 ± 12.2 years, and invasive breast carcinoma of no special type was the predominant histological subtype. The T-allele frequency in the study cohort was 0.50, whereas Latin American reference populations showed higher frequencies ranging from 0.69 to 0.83. RegulomeDB assigned rs5945619 a rank of 1f and a functional score of 0.22, supporting its regulatory potential. GTEx analysis identified tissue-specific cis-eQTL signals, including LINC01496 upregulation in testis, NUDT11 and LINC01496 downregulation in prostate, and modest GSPT2 downregulation in mammary tissue. This study provides the first characterization of rs5945619 in Ecuadorian mestizo women with BC and suggests its role as a potential X-linked regulatory variant, requiring further validation. Although universal heterozygosity was observed, the small sample size, tumor-only design, and absence of a healthy control cohort prevent any causal, risk-factor, or tumor-driven selection inference. The in silico evidence supports a tissue-dependent regulatory model involving LINC01496 and NUDT11, mainly in prostate/testis, and a modest GSPT2 signal in mammary tissue. Therefore, breast cancer-specific eQTL analyses, functional validation, and larger ancestry-informed case–control studies are required to determine the biological and clinical relevance of this locus. Full article

The Hasse derivative image of a repeated-root cyclic code is typically non-cyclic. This prevents a direct ideal-theoretic calculation of derivative-twisted intersections and first leads to cyclic containers and two-sided bounds. This paper shows that, after expanding codewords in Hasse–Taylor coordinates at the roots of the underlying polynomial, the higher-order Hasse derivative operator decomposes into independent local blocks and its image becomes a coordinate subspace. This yields exact closed-form dimensions, expressed entirely in terms of the generator multiplicities and the base-prime digits of the derivative order, for the derivative image, for the derivative-twisted intersection of two codes and its self-intersection specialization, and for the smallest cyclic code containing the image together with its non-cyclic defect. As a consequence, for every positive derivative order in the admissible range, the derivative image is cyclic only when it is zero, and the multiplicity-drop and Lucas-refined containers developed here, as well as the rank of the derivative operator, are recovered as immediate relaxations or special cases. Applying the intersection-pair construction to a code paired with its derivative image produces entanglement-assisted quantum error-correcting codes whose dimension and entanglement cost are exact functions of the multiplicity digits and the derivative order, so that the derivative order tunes the entanglement consumption. For the minimum distance, a punctured matrix-product decomposition gives an exact formula for the derivative image in terms of the minimum distances of a finite family of simple-root constituents of length $n_0$. Combining this formula with the corresponding constituent formula for the dual gives the complete distance parameter of the resulting EAQECCs without enumerating the non-cyclic derivative images. Full article

Background/Objectives: Ischemic heart disease (IHD) is a major cause of cardiovascular mortality, and emergency medical service (EMS) crews often make the first decision on whether an IHD-coded patient should be transported to hospital or managed at the scene. Evidence from Central Asian EMS systems is sparse, and it remains unclear whether routinely coded comorbidity information adds decision-relevant information beyond acute presentation. We aimed to identify predictors of the EMS hospitalization/transport decision among IHD calls in Astana, Kazakhstan. Materials and Methods: We conducted a retrospective call-level cohort study of 9985 consecutive EMS calls coded as IHD (ICD-10 I20-I25) over a five-year period. The endpoint was field disposition—hospitalization/transport versus being left at the scene—and should not be interpreted as confirmed ACS, mortality, clinical appropriateness or any other patient outcome. Group comparisons used the Mann–Whitney U and Pearson chi-square tests, and independent associations were estimated using explanatory multivariable logistic regression. Results: Overall, 2676 calls (26.8%) resulted in hospitalization/transport. The strongest independent predictors were cardiogenic shock (aOR 15.06), acute/unstable IHD versus chronic I25 (aOR 8.52) and heart failure (aOR 2.46). Other arrhythmias (aOR 1.84), atrial fibrillation (aOR 1.60), male sex (aOR 1.65) and age <45 years (aOR 1.88) were also associated with higher transport odds, whereas age ≥75 years (aOR 0.61), specialized crews (aOR 0.84) and high dispatch urgency (categories 1–2; aOR 0.84) were associated with lower odds. Model discrimination was moderate (AUC 0.69; optimism-corrected AUC 0.69), plausibly reflecting the absence of ECG findings, vital signs, symptom severity and hospital outcome data. Conclusions: The expected acuity markers dominated EMS transport decisions, but routinely coded cardiac comorbidities were independently associated with disposition in this understudied setting. The model is not deployable for individual triage; these variables should be considered candidate inputs for future models that incorporate richer clinical data, outcome linkage and prospective validation. Full article

Peptides are the most extensively studied bioactive compounds derived from food. They are analyzed using, e.g., in silico strategy. The BIOPEP-UWM database has become a standard tool in computer-aided peptide research. The aim of this study was to equip this database with a tool enabling the annotation and processing of peptide or protein sequences containing modified amino acid residues as well as other residues. The most recent section of the BIOPEP-UWM, i.e., the repository of amino acids and modifications, apart from 20 proteinogenic amino acids, annotates non-amino acid moieties or residues subjected to enzymatic or chemical modifications (phosphorylation, oxidation, hydroxylation, acylation, etc.). This part of BIOPEP-UWM provides the following information: Compound ID, name, symbol in a special code, InChIKey identifier, SMILES representation, number in the PubChem database (CID), formula, and IDs in other databases. The search options include ID in the repository, name, symbol in a biological code, InChIKey, PubChem Compound identifier (CID), and chemical formula. Peptide or protein sequences annotated using symbols from the repository are utilized by all applications available in the BIOPEP-UWM database. The database meets contemporary trends involving modifications of amino acid residues in the bioinformatic analysis of peptides and proteins. Full article

Perioperative stress induces profound molecular and cellular responses that contribute to postoperative complications, including perioperative neurocognitive disorders (PND), chronic postsurgical pain, organ dysfunction, immunothrombosis, fibrosis, and cancer progression. Increasing evidence demonstrates that epigenetic mechanisms act as central regulators linking surgical trauma, inflammation, metabolic stress, ischemia–reperfusion injury, and immune activation to long-term alterations in gene expression and tissue remodeling. DNA methylation, histone modifications, chromatin remodeling, non-coding RNAs, and RNA epitranscriptomic mechanisms such as N6-methyladenosine (m6A) collectively orchestrate perioperative responses across multiple organ systems. Recent translational studies have identified histone deacetylases (HDACs), histone methyltransferases, NETosis-associated chromatin signaling, HMGB1/NF-κB activation, and epigenetic regulation of neuroimmune pathways as major contributors to postoperative cognitive dysfunction, chronic pain, cardiac dysfunction, pulmonary injury, and fibrosis. In parallel, advances in liquid biopsy, circulating tumor DNA (ctDNA), and single-cell epigenomics have opened new opportunities for biomarker-guided perioperative precision medicine. This review summarizes current evidence regarding epigenetic regulation in perioperative medicine with special emphasis on neuroepigenetics, NETosis, fibrosis, cardiac epigenetics, immune remodeling, and perioperative oncological outcomes. Furthermore, we discuss emerging therapeutic strategies targeting HDACs, DNA methylation, m6A pathways, and chromatin-associated inflammatory signaling as potential future interventions for perioperative complications. Full article

Objectives: Oral potentially malignant disorders (OPMDs) carry a variable risk of malignant transformation, making early detection important. Deep learning relies on specialized imaging, which is often inaccessible in routine practice, and detection from standard clinical photographs remains poorly characterized. We assessed the feasibility of automated OPMD detection from white-light intraoral photographs, compared two convolutional neural network paradigms (global classification with DenseNet-121 versus one-stage instance segmentation with YOLOv8), and identified the main barriers to clinical translation. Methods: A dataset of 1500 photographs (750 OPMD and 750 non-OPMD) from institutional archives and publicly accessible sources was split in an 80:20 ratio for training and testing. Lesion boundaries were annotated by three trainees using Cytomine and validated by three specialists. The DenseNet-121 and YOLOv8-large-segmentation models were evaluated for accuracy, sensitivity, specificity, precision, F1 score, and Wilson 95% CI. Results: DenseNet-121 required extensive manual lesion cropping to converge, negating the automation rationale. YOLOv8-large-segmentation reached 62.2% accuracy (95% CI 56.4 to 67.6), 75% sensitivity (95% CI 67.3 to 81.4), 59.7% precision (95% CI 52.4 to 66.5), 49.3% specificity (95% CI 41.3 to 57.4), and an F1 score of 66.5%, detecting lesions in 96% of the test set. High sensitivity was obtained at a low confidence threshold of 0.1, with correspondingly reduced specificity, and the model produced interpretable color-coded segmentation masks. Conclusions: One-stage instance segmentation is the more viable direction and yields spatially interpretable output, but performance at this dataset scale is not yet clinically sufficient. Dataset scale, threshold calibration, low specificity, and absent patient metadata are the key barriers to address. Full article