Search Results (174)

Robust Arithmetic Word Problem (AWP) solving is important for applying mathematical reasoning systems in educational scenarios, where problem statements may contain changed numerical values, paraphrased descriptions, or irrelevant distracting information. Although Large Language Models (LLMs) have shown strong potential in solving AWPs, their reasoning processes may still be sensitive to surface-form variations and perturbation-induced noise. To address this issue, this paper proposes a Scene-Aware Neuro-Symbolic solver designed to improve the robustness of AWP solving under perturbations. The proposed method extends the existing scene-aware framework by introducing perturbation-oriented mechanisms at the scene, relation, and symbolic-solving levels. A Chain-of-Scene (CoS) prompting strategy first generates candidate scenes, after which goal-guided filtering retains target-related and bridge scenes while removing distractor-induced scenes. The retained scenes are then processed by the Scene-Aware Syntax-Semantics ($S^2$) method to extract explicit and implicit relations, and relation consistency checking is applied to remove locally plausible but globally irrelevant relations. Finally, the symbolic solver performs iterative equation-based reasoning over the filtered relation sets, with fallback recovery activated when standard solving does not produce a target-compatible answer. Experiments on AGG, MAWPS, and GSM8K show an average accuracy of 92.8% on clean datasets. On GSM-Perturb and AWP-Perturb, the solver achieves perturbed accuracies of 80.8% and 87.5%, with robustness drops of 8.3% and 6.8%, respectively. Ablation results show that scene filtering and relation consistency checking are the main contributors to reducing perturbation-induced errors. These findings suggest that combining LLM-based scene understanding with symbolic relation reasoning is a promising direction for improving the robustness and interpretability of AWP solvers in the evaluated perturbation settings. Full article

Background: Fragile X syndrome (FXS) is the most common monogenic cause of inherited intellectual disability and is primarily caused by CGG repeat expansion in the FMR1 gene. Conventional diagnostic methods have limited precision for sizing long repeat sequences and cannot resolve AGG interruptions, which are critical for comprehensive risk assessment. Existing national FXS reference materials are based on conventional methods and provide limited molecular information. Methods: We developed a targeted long-read sequencing assay for comprehensive FMR1 characterization, termed tLRS-FMR1, and applied it to a panel of 22 national FXS reference materials. Results: The tLRS-FMR1 assay demonstrated 100% concordance with standard methods while overcoming key limitations of conventional approaches. It enabled precise quantification of CGG repeat numbers, including full mutations (>200 repeats) that were only qualitatively reported by traditional techniques and provided comprehensive mapping of AGG interruption patterns. The assay showed high reproducibility, with 100% genotyping concordance across intra- and inter-assay replicates and achieved a detection limit of 3 ng/μL. Conclusions: This study successfully developed tLRS-FMR1 and established a new-generation national FXS reference material system with expanded molecular information and improved precision, providing a foundation for advancing the standardization and accuracy of FXS molecular diagnosis. Full article

Accurate contour extraction of aggregate particles from conveyor-belt depth maps is essential for downstream particle counting and size measurement, yet industrial depth data often contains weak discontinuities, missing values, and speckle-like noise. We propose a task-specific geometry-aware contour extraction framework that combines a compact SegFormer encoder with depth-derived priors, a lightweight local branch, edge-prior gated fusion, and full-resolution residual refinement. The input representation consists of normalized depth, Sobel gradient magnitude, and the absolute Laplacian response. On AGG_FULLDATA, the method achieves Optimal Dataset Scale (ODS), Optimal Image Scale (OIS), and Average Precision (AP) values of 0.9607/0.9716/0.9683 under the primary tolerance-based protocol ($tol=1$), while retaining an ODS of 0.6476 under strict pixel-exact matching. On External130, a test-only split collected under altered operating conditions using the same sensor, it reaches 0.9580/0.9734/0.9683 without retraining and consistently outperforms the MiT-only baseline. A rigid-object repeatability study based on 30 raw PLY scans shows a mean boundary deviation of 0.335 px, a within-1 px correspondence rate of 97.1%, and a coefficient of variation (CV) of equivalent diameter below 1%, supporting the practical meaning of $tol=1$. The full pipeline runs at 48.9 frames per second (FPS) with 3.71 M parameters on an NVIDIA GeForce RTX 4060 GPU. Broader robustness to separately controlled operating factors, environmental disturbances, and cross-device settings still requires validation. Full article

The occurrence of the European brown hare (Lepus europaeus Pallas 1778) has declined throughout Europe in recent years. However, it remains economically valuable as an important game species. To date, information on the individual microbiota of the European hare has been limited. The phylum Firmicutes (Bacillota) was dominant, and enterococci belong to this phylum. However, they can carry virulence factor genes. Therefore, this study aimed to address two aspects: the health of hares due to their recent decline, and, as a game animal, the protection of consumers’ health. Based on MALDI-TOF mass spectrometry, five strains were identified as Enterococcus faecium and two as E. faecalis; these findings were confirmed by genotyping using PCR and phenotypic analysis. The average value of lactic acid production was 0.680 ± 0.005 mmol/L. The strains lacked the virulence factor genes esp, agg, and gelE. However, they showed susceptibility to antibiotics and to postbiotic substances, even to 13 of 14 tested. PS/Ent M appears to be the most active PS against tested strains, with inhibitory activity of 25,600 AU/mL. Postbiotic substances represent a new tool for preventing unwanted microbiota in game animals. Full article

Most adversarial attack methods achieve high success rates under the white-box setting. However, these methods often lack transferability when targeting other deep neural network (DNN) models. Momentum-based attacks have emerged as an effective strategy to enhance transferability by incorporating a momentum term to stabilize update directions. While simple constant-momentum methods (e.g., MI-FGSM) or advanced variants (e.g., NI-FGSM, VMI-FGSM) have shown promise, they either use a single momentum decay factor or introduce significant computational overhead. To address this, we propose a novel plug-and-play momentum aggregation framework named AggMo-Attack. Our key insight is that a single momentum term with a fixed decay factor cannot optimally capture the multi-scale temporal correlations in gradients during adversarial optimization. Inspired by the Aggregated Momentum (AggMo) optimizer, we designed a multi-momentum aggregation module that maintains and weightedly combines multiple velocity vectors with different decay factors. This framework can be seamlessly integrated into existing momentum-based attack methods (e.g., MI-FGSM, NI-FGSM, VMI-FGSM) as a drop-in replacement for their standard momentum update step. Extensive experiments demonstrate that integrating our AggMo module significantly improves adversarial transferability. Our work provides a versatile and effective tool for enhancing momentum-based adversarial attacks and opens a new direction for designing adaptive attack strategies. Full article

This study was conducted in Northwest Ethiopia in 2025 to estimate genetic parameters for dairy cattle hybrids under a group-based mass selection scheme. The objective was to investigate lactation milk yield (MY), lactation length (LL), and key fitness traits across varying breed compositions, aligned with suitable agro-ecological zones and milkshed systems. The findings may then serve as a framework to develop economically efficient and sustainable dairy genotypes tailored to the region. Data were collected from 355 dairy households using semi-structured questionnaires and monthly monitoring of MY. A mass selection scheme was applied to evaluate the productive and reproductive performance of Holstein-Friesian (HF) and Jersey hybrids across varying levels of exotic breed compositions. To identify superior genotypes, a total merit index (TMI) was developed, utilizing economic weights of +0.20 for production traits and −0.12 for reproductive traits. General liner model (GLM) analyses were performed to evaluate the performance of different breeds and exotic breed composition. Realized genetic parameters including genetic correlations (rg) as an indicator of pleiotropy, genetic gain (GG) per trait, and aggregate genetic response (AGG) were estimated for each group using specialized procedures in R software. Breed type (stratified by exotic breed composition), agro-ecology zone, and milkshed system were defined as the main and sub-fixed effects. The genetic contribution to the performance of hybrids indicated that the Holstein-Friesian (HF) hybrid baseline scheme achieved significantly higher efficiency, with an aggregate genetic gain) (AGG) of 155.50, compared with 136.03 for the Jersey hybrid schemes. Specifically, the >75% HF hybrid group exhibited the highest predicted AGG (183.00), a result primarily underpinned by significant gains in MY (182.53 L) and extended LL (0.28 months). This indicated that higher exotic breed composition in HF crosses maximizes the genetic gain when selection is weighted toward productivity. Conversely, the 62.5% Jersey hybrid exhibited the lowest AGG (110.38) and GG for MY (109.86 L), indicating that intermediate Jersey breed compositions may be suboptimal under the studied conditions. Analysis of interaction effects revealed environment-specific superiorities: in the Bahir Dar midland milkshed, the >75% HF hybrids achieved the highest genetic gains in MY (182.53 L) and a superior AGG (181.34). In contrast, within the Gondar midland milkshed, >75% Jersey hybrids reached the highest overall AGG (177.11), with a corresponding GG for MY of 178.75 L per lactation. The observed variance in MY (δ² = 362.44) indicated significant potential for genetic improvement through group-based selection. Pleiotropy was identified between MY and LL (r_g = 0.14), whereas an antagonistic trade-off was observed between maturity and conception efficiency (r_g = −0.34). The consistent upward trend in the performance of hybrids as breed composition increased from 50% to >75% across both main and sub-effects suggests that these genotypes are suited to the environment. In conclusion, single- and multiple-trait predictions based solely on breed and breed comparisons were suboptimal; instead, selection strategies incorporating genotype-by-environment (G × E) interactions offered the most effective alternative for regional dairy selection alternatives. Full article

Epidermal Growth Factor (EGF) plays an important role in skin regeneration and repair by promoting cell proliferation and collagen synthesis. However, its topical application is limited by low stability, susceptibility to degradation, and poor penetration through the stratum corneum due to its hydrophilic nature and relatively large molecular size. Microencapsulation offers a strategy to protect sensitive bioactives and improve their delivery in cosmetic formulations. In this study, EGF was encapsulated in Arabic gum/gelatine A (AG/GE) coacervate microcapsules and incorporated into a hydrating cream. The work extends previous studies using the same microcapsule composition for lipophilic compounds, demonstrating its applicability for a hydrophilic bioactive and highlighting the versatility of the encapsulation platform. The resulting microcapsules exhibited spherical, multinucleated morphology with an encapsulation efficiency of 78.8 + 1.0%. Although diffusion of microencapsulated EGF in the cream could not be directly determined, the formulation showed trends towards improvement in several skin parameters during the volunteer evaluation, including reduction in surface spots (31%), brown spots (21%) and pore visibility (10%), and improved texture (22%). A 25% decrease in transepidermal water loss and a 33% increase in elasticity suggested improved skin barrier function. Volunteers reported high acceptance regarding non-irritancy, texture, and sensory experience. Full article

Secure and bandwidth-conscious transmission of model updates is a central bottleneck in distributed machine learning. Existing secure aggregation and homomorphic encryption pipelines either reveal more than the task requires or incur prohibitive computation and communication costs. We introduce a verifiable functional encryption (VFE) framework that releases only the intended linear functions of client gradients while providing end-to-end integrity and privacy guarantees under standard lattice assumptions. Our instantiation, FlowAgg-FE, combines two novel components. First, KS-IPFE, a key-splittable inner-product FE scheme, supports per-round weighted aggregation, vector packing, and on-the-fly function changes without client re-encryption; function keys are distributed across two non-colluding helpers, eliminating a single point of trust and enabling lightweight, homomorphically verifiable tags on decrypted outputs. Second, PaS-Stream is a rate-adaptive encryption-and-compression pipeline that couples sketch-based gradient compression with batched FE ciphertext streaming, ensuring unbiased aggregation in the presence of stragglers and dropouts. We further bind client-side clipping to zero-knowledge range proofs and offer an optional differentially private release layer that composes with FE to yield $(\epsilon ,\delta )$-privacy. A prototype based on LWE demonstrates practicality across cross-device and cross-silo training: client uplink is reduced by 1.9–3.4× and server CPU time by $1.6\times$ versus state-of-practice encrypted secure aggregation, with accuracy within $0.3\%$ of plaintext baselines and correctness preserved under up to $30\%$ client dropout. These results show that verifiable FE can make secure, communication-efficient gradient transmission viable, as appropriate for theme of security and privacy in distributed machine learning of the Special Issue. Full article

Background/Objectives: As the prevalence of depression and the use of antidepressants have risen steadily in the last decade, new treatment options are needed. Aloysia gratissima var. gratissima ethanol extract has previously shown antidepressant-like activity, and the present study was conducted to identify the active fraction and clarify the possible mechanisms of action. Methods: Tail suspension (TST) and forced swimming (FST) behavioral tests were performed, and possible mechanisms of action were elucidated using serotonergic, dopaminergic, adrenergic, and GABAergic system antagonists. UPLC-DAD-MS analyses were performed to identify compounds in active fractions, and molecular docking studies were carried out to determine the binding affinities of these compounds to serotonergic and dopaminergic receptors (5-HT_1A, 5-HT_2A, 5-HT₃, and D₂R). Results: Ethyl acetate and butanol fractions were found to decrease immobility time in FST. The reduction in immobility time during the FST caused by the ethyl acetate fraction was reversed by pretreating mice with WAY100635 (5-HT_1A antagonist), ketanserin (a 5-HT_2A antagonist, ondansetron (5-HT₃ antagonist), or haloperidol (D₂ antagonist). UPLC-DAD-MS analysis revealed a similar composition for the ethyl acetate and butanol fractions of A. gratissima var. gratissima. Pharmacokinetic predictions suggest that only a few of the identified compounds have the potential to permeate the blood–brain barrier, and molecular docking simulations showed that compounds such as 13-oxooctadecadienoic acid, ferulic acid, and coumaric acid have binding affinities to the druggable site of serotonergic and dopaminergic receptors. Conclusions: These results suggest that the Agg ethyl acetate fraction possesses antidepressant-like activities, altering dopaminergic and serotonergic system functions. Computational simulations also suggest that some of the identified compounds have binding affinities to the 5-HT_1A, 5-HT_2A, 5-HT₃, and D₂R receptors. Full article

Early and rapid diagnosis of drug resistance in tuberculosis (TB) plays a key role in reducing the spread of resistance and enabling effective treatment. The aim of this study was to investigate mutations in drug resistance-associated gene regions of Mycobacterium tuberculosis complex (MTBC) isolates resistant to at least two first-line anti-tuberculosis drugs through sequence analysis, in order to characterize the core molecular features of these strains in the region and to identify previously unreported, geographically distinct novel mutation sites. The drug susceptibility of 23 clinical isolates was assessed using the BACTEC MGIT 960 system, and resistance-associated point mutations were identified through DNA sequence analysis and comparison with GenBank reference sequences. AAG → AGG mutation was detected in the rpsL gene region at codon 43 (n = 7) and codon 88 (n = 1). Additionally, GAG → GCG point mutation was identified at codon 70 (n = 2), representing a new region not previously reported in the literature. The most frequent mutation was AGC → ACC at katG codon 315 (n = 10), followed by a C → T substitution at position −15 of the inhA promoter region (n = 4). Additionally, TCG → TTG at rpoB codon 531 (n = 4) and ATG → GTG at embB codon 306 (n = 1) were detected. The detection of resistance-associated mutations is essential for controlling drug-resistant tuberculosis. In this study, a novel rpsL mutation (GAG → GCG) at codon 70 and a previously unreported codon 88 mutation in our country were identified, contributing to the understanding of molecular resistance mechanisms and epidemiology. Full article

Background: The Phoridae are one of the most poorly studied families of Diptera insects in Russia. They are small flies that play an important role in ecosystems. Methods: This dataset presents the results of a study on Phoridae conducted between 2019 and 2024 in European Russia. The overall study area covered 400,000 km². Results: A total of 16,265 specimens were reliably identified, representing 272 species and 22 genera from 180 localities. Of these, 2673 specimens were females (16.4%), while the remaining 83.6% were males. Conclusions: The genus Megaselia Rondani accounted for 200 species (73.5%) and 12,120 specimens (74.5%). Ten species were particularly common: Megaselia pusilla, M. angusta agg., Triphleba opaca, Diplonevra funebris, M. brevicostalis, M. plurispinulosa, M. flavicans, M. lutea, M. minuta, and M. lactipennis. The highest number of localities was recorded for M. angusta agg. (37.2%), M. flavicans (27.8%), and M. brevicostalis (25.0%). In terms of collection methods, the majority of both specimens and species were captured using Malaise traps and pan traps. The highest species richness and specimen abundance were recorded in floodplain habitats, steppified areas, and meadows. In contrast, forested sites showed lower species diversity and abundance. Full article

Enterococci from raw goat milk were taxonomically allotted in the species Enterococcus faecalis using sequencing (16S rRNA and BLASTn analysis) with a percentage identity up to 99.91%. The virulence factor gene gelE was found in the strains EE/K3, EE/G3, and EE/G6. The agg gene was detected in the strain EE/G6, and the esp gene was detected in the strains EE/K5 and EE/G7. Each strain possessed at least one virulence factor gene. In the strain EE/G6, the gelE and esp genes were found. The strains EE/G6 and EE/G3 showed resistance to tetracycline and vancomycin. EE/G7 was resistant to vancomycin and gentamicin. All strains possessed low-grade biofilm-forming ability (0.1 < A570 ≤ 1.0). They possessed genes for biofilm formation (bopD, srt, and/or ace). They also produced esterase (20–40 nmo/L), esterase lipase, and α-chymotrypsin (10–40 nmoL). The values of acid phosphatase reached 20–40 nmoL. The strains EE/G3, EE/G6, and EE/G7 were observed to possess the most pathogenicity. However, all strains were susceptible to postbiotic active substances produced by two autochthonous lactococci, MK2/8 and MK1/3 (inhibitory activity up to 400 AU/mL). These postbiotic substances provide a new potential alternative to reducing contaminants in milk. Full article

We conducted a comparative analysis of the identified microsatellite sequences across the genomes of nine sillaginids. We examined the microsatellites with motifs ranging from 1 to 6 bp in length and analyzed their distribution and frequency across different genomic regions. Microsatellite occurrence differed significantly with the degree of coverage ranging from 1.47 to 3.21%. The number and proportion of each repeat type were consistent across the nine species, with di-nucleotide repeats being the most abundant, followed by mono-nucleotide repeats, and gradually decreasing as the number of repeat units increased. The mono-nucleotide repeat motifs were dominated by A/T, while di-nucleotide repeat motifs were mainly AC/GT, and tri-nucleotide repeat motifs were primarily AGG/CCT. Regarding the number of repeats, microsatellites in Sillaginidae were generally concentrated between 5 and 18 repeat units, with peaks observed at 6 and 10 repetitions. The abundance of microsatellite loci consistently decreased as the number of repetitions increased beyond 10. These findings provide valuable insights into genome evolution and microsatellite DNA dynamics, supporting future investigations into their structural and functional characteristics, compositional patterns, and applications in molecular marker development for these species. Full article

The fungal genus Trichoderma is highly valued in agriculture for its versatile roles, mainly as a biocontrol agent against plant pathogens. Recently, its use as a natural biofertilizer has gained attention, as Trichoderma spp. promotes crop growth and improves yield by enhancing the rhizosphere environment and activating plant defences. Globally, over 250 Trichoderma-based products dominate 60–90% of the market, but their efficacy can decline during transportation and storage. Additionally, concerns about their impact on native soil biodiversity have led to interest in using locally adapted, native strains. The novel native strain of Trichoderma sp. STP8 (formerly T. koningiopsis agg. STP8) previously showed strong antagonism against Sclerotinia sclerotiorum and promoted lettuce growth in greenhouse conditions. This study evaluated Trichoderma sp. STP8’s effectiveness in field-grown lettuce, revealing yield increases of 16.6% to 30.5%. The most significant gains occurred when Trichoderma sp. STP8 was applied before head formation, 26 days after planting. That was in one treatment with two applications (at seedling planting and after 26 days) and another with three applications (at sowing, at seedling planting, and after 26 days). These results demonstrate Trichoderma sp. STP8’s potential as a sustainable biocontrol and biofertilizer agent for lettuce, encouraging further research across different agricultural systems, including hydroponics and soil-less medium. Full article

Image blur is a major factor that degrades object detection in agricultural applications, particularly in orchards where crop occlusion, leaf movement, and camera shake frequently reduce image quality. This study proposed a lightweight generative adversarial network, AGG-DeblurGAN, to address non-uniform motion blur in citrus tree images. The model integrates the GhostNet backbone, attention-enhanced Ghost modules, and a Gated Half Instance Normalization Module. A blur detection mechanism enabled dynamic routing, reducing computation on sharp images. Experiments on a citrus dataset showed that AGG-DeblurGAN maintained restoration quality while improving efficiency. For object detection, restored citrus images achieved an 86.4% improvement in mAP@0.5:0.95, a 76.9% gain in recall, and a 40.1% increase in F1 score compared to blurred images, while the false negative rate dropped by 63.9%. These results indicate that AGG-DeblurGAN can serve as a reference for improving image preprocessing and detection performance in agricultural vision systems. Full article