Browse Articles

A robust multi-step predictor-corrector method for solving fractional order biological models

Disaggregating between- and within- person effects of ovarian hormones on the association between worry and error-related brain activity in a female sample

Expression of Concern: MiR-125b Reduces Porcine Reproductive and Respiratory Syndrome Virus Replication by Negatively Regulating the NF-κB Pathway

Exploring agro-morphological diversity and superior genotypes in Beta vulgaris var. bengalensis L.: a two-year comprehensive study of an underutilized traditional green leafy vegetable of Northern India

In vivo rectal dosimeter with MRI marker

Background Hypofractionated external beam radiotherapy for prostate cancer necessitates precise rectal dose evaluation. We fabricated a radiochromic polyurethane-based in vivo rectal dosimeter with a custom MRI marker and a patient-specific applicator for in vivo dose verification (IDV) during gated MR-image guided radiotherapy (MR-IGRT). Methods The dosimeter featured a radiochromic polyurethane active layer, incorporating leucomalachite green (LMG) and tartrazine. To accommodate anatomical variations, a detachable PMMA applicator was designed in five sizes. For localization, four elastomeric materials, including two polyurethane-based and two silicone-based materials, were evaluated as candidate MRI markers. Post-irradiation fading was assessed over time to evaluate measurement stability. A dose-response calibration was performed to establish a linear relationship between net optical density (OD) and absorbed dose. Furthermore, dose uncertainty was analyzed based on the law of error propagation. For verification, in vivo measurements were conducted for two patients and compared with TPS-calculated doses. Results Vyta Flex 20 was selected as the optimal MRI marker due to its high signal intensity and ease of fabrication. Post-irradiation net OD showed a dose-dependent temporal response, and the readout time was standardized to 2 h. The dosimeter’s sensitivity was 0.00253 cGy -1 . Dose uncertainties were determined to be 2.1%, 2.0%, and 1.6% at 100, 200, and 300 cGy, respectively. In vivo verification showed mean dose differences of 3.7 ± 1.4% (95% CI, 0.1–7.3%) for patient #1 and 5.9 ± 2.1% (95% CI, 0.8–11.1%) for patient #2. Measured doses were consistently higher than TPS-calculated doses, suggesting possible contributions from localization uncertainty in high-dose-gradient regions and the material-dependent response of the radiochromic active layer. Conclusion The fabricated radiochromic dosimeter with a custom MRI marker and adjustable applicator demonstrated preliminary feasibility as a proof-of-concept system for in vivo rectal dose verification during MR-IGRT. Further studies with larger patient cohorts and more treatment fractions are required to validate its reproducibility, statistical robustness, and clinical utility. Further refinement in positioning is also needed, particularly in dose-gradient regions.

Discriminative performance of serum homocysteine in drug-naïve depression: a case-control study with ROC and regression analysis

Abstract Depression is a multifactorial disorder involving neurochemical, metabolic, and inflammatory dysregulation. Elevated homocysteine, a marker of impaired one-carbon metabolism, has been implicated in neurotoxicity and oxidative stress pathways associated with depressive pathology. This study evaluated serum homocysteine levels and their association with body mass index (BMI) in patients with incident drug naiive depression compared to healthy controls. This case-control study enrolled 200 participants: 100 patients with newly diagnosed depressive disorder (DSM-5 criteria) and 100 age- and sex-matched healthy controls recruited from a tertiary psychiatry centre in South India. Serum homocysteine was measured using standard enzymatic assay and BMI was calculated from standardised anthropometric measurements. Independent samples t-test, Pearson’s correlation, ROC analysis and Binomial logistic regression analyses were performed. Serum homocysteine was significantly higher in depressed patients than controls (35.8 ± 7.04 µmol/L vs. 13.6 ± 2.48 µmol/L; p  < 0.001, Cohen’s d = 4.193; 95% CI for difference: 20.66–23.58 µmol/L). BMI was also significantly elevated in the depression group (25.67 ± 4.97 kg/m 2 vs. 21.92 ± 2.96 kg/m 2 . p  < 0.001, Cohen’s d = 0.912). A positive homocysteine—BMI correlation was observed in depression( r  = 0.42, p  < 0.001) but not in controls ( r  = 0.04, p  = 0.71). ROC analysis yielded an AUC of 1.000 (optimal cutoff: 20.065 µmol/L, sensitivity: 100%, specificity: 100%), reflecting complete distributional separation in this clinical sample. Binary logistic regression analysis showed homocysteine to be a significant predictor of depression(omnibus χ 2 = 277.26, df = 1, p  < 0.001; 100% classification accuracy; Nagelkerke R 2 = 1.00), though standard odds ratio estimation was precluded by complete separation. Elevated homocysteine and a group-specific BMI—homocysteine correlation were identified in depressed patients. The AUC of 1.000 reflects complete distributional separation in a tightly controlled institutional sample and requires external validation before clinical application. As folate, vitamin B12, and lifestyle confounders were unmeasured, findings are hypothesis-generating and warrant prospective multicentre validation.

Evaluation of analgesic and anti-inflammatory activities of the root extract of Grewia schweinfurthii Burret and its major chemical constituents

Despite the widespread use of conventional analgesic and anti-inflammatory drugs, their clinical utility is often limited by adverse effects, necessitating the search for safer alternatives from medicinal plants. This study investigated the analgesic and anti-inflammatory activities of the 80% methanolic root extract of Grewia schweinfurthii Burret and an isolated bioactive compound using established rodent models. Air-dried roots were extracted by maceration with 80% methanol, and acute oral toxicity was evaluated following OECD guidelines. Analgesic activity was assessed using the acetic acid–induced writhing and hot plate models in mice, while anti-inflammatory activity was evaluated using carrageenan-induced paw edema in rats. A bioactive compound, 4 (2ʺ-(4′-isopropylphenyl) propan-2ʺ-yl)-2,3-dihydrofuran, was isolated via column chromatography and tested for anti-inflammatory activity. In the writhing test, the extract produced significant (p < 0.001) dose-dependent inhibition of abdominal constrictions, with 13.90%, 56.81%, and 75.48% inhibition at 100, 200, and 400 mg/kg, respectively, compared to 80.77% inhibition by aspirin (150 mg/kg). In the hot plate model, the 400 mg/kg dose significantly prolonged latency time from a baseline of 5.67 ± 0.33 s to 9.17 ± 1.01 s at 120 min (p < 0.05), indicating central analgesic activity. In the carrageenan-induced paw edema model, the extract demonstrated marked anti-inflammatory effects, with 93% inhibition at 400 mg/kg at the 5th hour, comparable to indomethacin (95% inhibition). The isolated compound exhibited significant dose-dependent anti-inflammatory activity, achieving 75% inhibition at 40 mg/kg at 5 hours (p < 0.001). Overall, the findings demonstrate that G. schweinfurthii root extract and its isolated compound possess significant peripheral and central analgesic as well as potent anti-inflammatory activities, supporting the plant’s traditional use; however, further mechanistic, toxicological, and pharmacokinetic studies are required before clinical relevance can be established.

Development and validation of a predictive AI model for differential diagnosis of endodontic and non-endodontic orofacial pain: a comparative study

Predictive value of complete blood cell count-based inflammatory markers for peritonitis risk in peritoneal dialysis patients: A multicenter cohort study

Background Peritonitis is a serious complication of peritoneal dialysis (PD). Inflammatory indices derived from routine complete blood count (CBC) parameters—including the pan-immune inflammatory value (PIV), systemic immune-inflammatory index (SII), platelet-to-lymphocyte ratio (PLR), neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), and platelet-to-monocyte ratio (PMR)—have shown prognostic value in various diseases. However, their comparative utility in predicting PD-associated peritonitis (PDAP) remains unclear. This multicenter cohort study aimed to evaluate and compare these indices to identify the best predictor of PDAP. Methods We retrospectively enrolled 2,036 PD patients from 10 centers. The associations between inflammatory markers (PIV, SII, PLR, NLR, MLR, PMR) and peritonitis risk were analyzed using restricted cubic splines. Optimal cut-offs were determined by ROC analysis. Survival differences were assessed using Kaplan–Meier curves and log-rank tests. Independent predictors were identified via multivariate Cox regression, with model discrimination evaluated by the C-index. Subgroup analyses were conducted by gender, age, body mass index (BMI), diabetes, albumin, PD vintage, and residual renal function. Results The median age was 51.0 years, 55.01% were male, and median dialysis vintage was 49.47 months. Diabetes prevalence was 21.02%. Over the follow-up, 147 patients (7.22%) developed peritonitis. Among the indices evaluated, PIV, SII, and PLR showed significant nonlinear associations with peritonitis risk (all P  < 0.05). Adjusted hazard ratios were 2.004 for PIV, 2.144 for SII, and 2.063 for PLR. Adjusted C-indices were 0.67 (PIV), 0.70 (SII), and 0.70 (PLR). No significant interactions were found in subgroup analyses. Conclusion Elevated PIV, SII, and PLR levels at PD initiation independently predict higher peritonitis risk. Although their discriminative ability is moderate, these routine, cost-effective indices may aid risk stratification and help identify patients needing closer monitoring or preventive interventions.

Microscopic pore characteristics governing macroscopic freeze thaw durability and service life prediction of ternary blended concrete

Preliminary study on comparative non-targeted metabolomics analysis sheds light on the chemical diversity of citrus fruit pulps

Citrus flavor and nutritional quality are closely tied to metabolite composition, yet comparative metabolic dissection of fruit pulp traits across citrus subspecies remains insufficient. Here, we applied non-targeted LC–MS/MS metabolomics to examine chemical diversity in the fruit pulp of three representative citrus varieties: Citrus reticulata ‘Hongju 418’, Citrus aurantium ‘Changshan-huyou,’ and Citrus junos ‘Hunan Xiangcheng.’ Through differential metabolite analysis, multivariate modeling, correlation network construction, and KEGG pathway enrichment, we characterized the extent and nature of metabolic divergence among these genotypes. PCA, PLS-DA, and OPLS-DA revealed distinct metabolic clusters, underscoring strong genotype-specific variation. More than 300 differentially expressed metabolites were identified, including flavonoid glycosides, organic acids, phenolic derivatives, and limonoids. Hongju 418 was enriched in flavonoid biosynthetic pathways, Xiangcheng accumulated higher concentrations of organic and amino acids, and Huyou displayed a unique hybrid profile marked by elevated fatty acid and purine metabolism. Correlation and KEGG analyses consolidated these observations, revealing coordinated pathway-level shifts that define subspecies-specific metabolic architectures. Collectively, this work deepens current understanding of citrus pulp chemotypes and provides a robust biochemical foundation for advances in citrus breeding, quality assessment, and functional product innovation.

Thermodynamic design and energy–exergy analysis of a water-cooled photovoltaic–thermal systems under tropical conditions

AgriOptNet: A hybrid optimization and lightweight deep learning framework for soil texture classification and crop recommendation based on nutrition

Agriculture is a central part of human subsistence, with classification of soil texture and nutrition-based crop recommendation being the central aspects of optimal agricultural practice. Nevertheless, traditional methods are subject to limitations of being less precise, computationally less optimal, and less versatile concerning varying soil and environmental conditions. Current deep learning models are frequently unable to compromise between performance and efficiency, whereas traditional optimization methods fail to handle high-dimensional agriculture data efficiently, resulting in suboptimal suggestions and poor real-time usage. To address these issues, this research presents AgriOptNet, a hybrid deep learning and optimization framework for intelligent soil texture classification and crop recommendation based on nutrition. AgriOptNet novelty is founded upon three integral constituents like Crop Recommendation through Entropy-Regularized Dynamic Deep Q-Learning with Adaptation to the Reward Function (MDQL-RA), optimally dynamic recommendations of crop inputs depending upon the health of soil, yield records, and surrounding environmental aspects and utilizing entropy regularization to accelerate exploration; Classification using a newly invented lightweight deep-learning model called SoilCropNet with a compound based on MobileNetV2, EfficientNetV2, and ShuffleNetV2 and provides precise, and computationally favourable classification along with squeeze-and-excitation as well as depth-wise separable convolutional enhanced properties; Feature selection through newly developed hybrid SailDragon Optimizer (SDO), combining Sailfish Optimization (SFOA) and Dragonfly-Based Optimization (DBOA), to obtain best-informing features for predictions without errors. The proposed AgriOptNet framework demonstrates superior performance with an accuracy of 99.87% and an F1-score of 98.75%, significantly outperforming existing techniques and ensuring high precision and efficiency for real-time precision agriculture applications.

Comparative analysis of deep learning models for garment segmentation in textile recycling

Abstract The textile industry is a key player in the global economy, but its expansion and resource consumption raise sustainability concerns. Addressing these issues is critical to fostering a more sustainable future. One key challenge in garment recycling is the removal of accessories and inserts before fibers can be recovered. This study presents a comparative analysis of deep learning models for garment segmentation as an enabling step for future automated textile recycling workflows. Accurate garment segmentation can support subsequent tasks such as garment classification, accessory localization, and selective removal, improving the consistency and efficiency of pre-processing before fiber recovery. Several state-of-the-art models, including You Only Look Once version 8 (YOLOv8), Inception Residual Network version 2 with U-Net (InceptionResNetV2-UNet), U2-Net, and Mask Region-based Convolutional Neural Network (Mask R-CNN), were evaluated to understand their strengths and weaknesses in handling garment complexity. The highest performance was demonstrated by the combination of the YOLOv8n detector plus U2-Net, with 99.56% accuracy, 99.11% F1-score, 98.23% Jaccard Index, 99.91% recall, and 98.33% precision, when tested on 347 laboratory images, showcasing robust segmentation capabilities under controlled conditions. The results suggest that this combination is well-suited for garment segmentation tasks and can provide a useful computer vision foundation for future accessory localization and automated removal modules. This comparison serves as a basis for future studies focused on improving segmentation robustness and integrating garment segmentation with downstream textile recycling processes.

Beyond citation-based metrics: Measuring interdisciplinarity via SBERT semantic embeddings and its heterogeneous effects on citation impact

Background Interdisciplinary research is a cornerstone of global science policy, yet decades of research have reached conflicting conclusions about its association with citation impact. This inconsistency stems primarily from traditional indicators, which rely on reference diversity rather than genuine semantic knowledge integration, and from small, discipline-specific samples that limit generalizability. Objective This study introduces a novel semantic interdisciplinarity measure based on Sentence-BERT (SBERT) embeddings, which directly captures cross-disciplinary knowledge integration at the textual level, and tests its heterogeneous relationship with citation impact across the full spectrum of scientific disciplines. Methods We analyzed 121,194 articles published 2015–2025 across all 19 root-level disciplines in OpenAlex. We validated the reliability of OpenAlex disciplinary classification using multi-dimensional semantic analyses, and compared our SBERT-based indicator with the Simpson Diversity Index and Rao–Stirling Index. We employed OLS and negative binomial regressions with discipline and year fixed effects (standard errors clustered at the discipline level), journal tier heterogeneity analysis, and domain-specific decomposition analyses. Results The semantic interdisciplinarity indicator shows moderate convergent validity with conventional citation-based metrics (r = 0.333–0.347, p < 0.001) and provides a small but meaningful increase in explanatory power beyond traditional indicators (Δ adjusted R² = 0.003), although its coefficient is marginally significant and negative (β = −1.5565, p = 0.085). Overall, semantic interdisciplinarity is positively associated with citation impact in baseline models, but this effect is primarily driven by cross-domain integration between epistemically distant domains, particularly in the natural sciences. The positive effect is consistent across all journal influence tiers, with the strongest effect observed in mid-tier journals, and presents stark heterogeneity across individual disciplines. Conclusion Boundary-spanning research bridging epistemically distant domains appears to deliver consistent citation rewards. Our findings address long-standing inconsistencies in the literature, and provide actionable insights for research evaluation and science policy.

Signal processing and machine learning analysis of IMU-based lower-limb kinematics in spinal sagittal imbalance

Characterisation of the bacterial and archaeal microbiota in processed colostrum collected from a spring-calving dairy herd

Colostrum feeding is critical for neonatal calf health, providing immunoglobulins (Ig) and other bioactive compounds that support immune function and early microbiome development. While the microbiota of fresh colostrum has been characterised, colostrum on commercial dairy farms is often refrigerated and reheated prior to feeding – practices that may alter its microbial composition. Therefore, the objective of this study was to characterise the prokaryotic community of refrigerated and reheated (processed) colostrum collected immediately before calf feeding. Twenty-one processed colostrum samples were collected from a single, primi- and multiparous Holstein-Friesian and Jersey, spring-calving dairy herd with no more than two donors contributing to each sample. Colostrum samples were refrigerated for no more than 24h and then re-heated in a 38°C water bath for 60 minutes. Colostrum samples were collected immediately prior to being fed to the calf. Microbial DNA was extracted and16S rRNA gene amplicon libraries were sequenced using the Illumina platform. Raw sequencing data were processed in R via the DADA2 pipeline, and an amplicon sequence variant (ASV) table was generated. Taxonomy was assigned using the SILVA database (v. 138.1) and data were subjected to α- and β-diversity analysis using Phyloseq, Microbiome and Vegan . Breed and parity had no effect (P ≥ 0.05) on α- and β-diversity. The mean Shannon index score (α-diversity) was 2.26 (SE 0.18), indicating unevenness and low levels of richness. Microbial community composition varied considerably between samples. Five archaeal ASV genus groups were identified, with Methanobrevibacter dominating this community(relative abundance (RA) of 85.19%). Four bacterial phyla were identified as the major contributors to the bacterial component of processed colostrum. Only 39 ASV genus groups were identified as having a RA > 0.05%. Processed colostrum was dominated by Pseudomonas (RA = 20.97%) and Acinetobacter (RA = 18.65%). These genera, along with 11 others, including Romboutsia, Flavobacterium. Lachnospiraceae NK3A20 group and Clostridium sensu stricto 1 were present across all samples and thus considered core bacteria. Overall, these findings indicate that refrigeration and reheating may significantly alter the natural colostrum microbiota, reduce diversity and increase heterogeneity of the community composition. Further research is needed to determine how these changes influence microbial seeding and calf health outcomes.

Whole-genome analysis of drug resistance and transmission patterns in Mycobacterium tuberculosis from Khyber Pakhtunkhwa, Pakistan

Abstract The emergence of drug-resistant Mycobacterium tuberculosis strains presents a significant challenge to tuberculosis (TB) control programmes worldwide, particularly in high-burden countries, such as Pakistan. This study characterises the genomic diversity and drug resistance mutation profiles of M. tuberculosis isolates from Khyber Pakhtunkhwa using whole-genome sequencing (WGS). A total of 35 clinical M. tuberculosis isolates underwent WGS. Sequence data were quality-filtered and mapped to the H37Rv reference genome, characterising variants, including single-nucleotide polymorphisms (SNPs). Lineage classification, drug resistance profiling, and identification of resistance mutations were performed using TB-Profiler software. SNP-based clustering was used to infer recent transmission events. All four major  M. tuberculosis  lineages (L1–L4) were detected; however, more than half of the isolates belonged to lineage L3 (CAS, n  = 19; 54.3%). Almost all isolates were drug-resistant (34/35), including pre-extensively drug-resistant (pre-XDR,  n  = 15; 42.9%) and multidrug-resistant (MDR,  n  = 14; 40.0%) strains. The most frequent resistance-associated mutations included  katG S315T (27/35) and  inhA c.-777C > T (4/35) for isoniazid resistance,  rpoB S450L (24/35) for rifampicin resistance,  embB M306I/V (18/35) for ethambutol resistance, and  gyrA D94G (8/35) and A90V (5/35) for fluoroquinolone resistance. Compensatory mutations were observed in  rpoC and  ahpC . Mutations associated with bedaquiline and clofazimine resistance ( mmpR5 c.140dupA, c.138_139dupTG) were detected in single isolates. Interestingly, one isolate exhibited a deletion encompassing  mmpR5, mmpL5 , and mmpS5 , potentially increasing bedaquiline susceptibility. Phylogenetic analysis using 6,635 genome-wide SNPs identified two transmission clusters with < 9 SNPs differences, suggesting ongoing local transmission. These results highlight the presence of drug-resistant  M. tuberculosis  strains in Khyber Pakhtunkhwa, encompassing both globally recognised and novel resistance-associated mutations. Our findings underscore the critical need to integrate WGS into TB surveillance and control programmes in Pakistan to guide targeted interventions and curb the spread of resistant strains.

Machine learning of electronic structure and atomistic properties from the external potential

Electronic structure calculations remain a major bottleneck in atomistic simulations and, not surprisingly, have attracted significant attention in machine learning (ML). Most existing approaches learn a direct map from molecular geometries, typically represented as graphs or encoded local environments, to molecular properties or use ML as a surrogate for electronic structure theory by targeting quantities, such as Fock or density matrices expressed in an atomic orbital (AO) basis. Inspired by the Hohenberg–Kohn theorem, in this work, we propose an operator-centric framework in which the external (nuclear) potential, expressed in an AO basis, serves as the model input. From this operator, we construct hierarchical, body-ordered representations of atomic configurations that closely mirror the principles underlying several popular atom-centered descriptors. At the same time, the matrix-valued nature of the external potential provides a natural connection to equivariant message-passing neural networks. In particular, we show that successive products of the external potential provide a scalable route to equivariant message passing and enable an efficient description of nonlocal effects. We demonstrate that this approach can be used to model molecular properties, such as energies and dipole moments, from the external potential or to learn effective operator-to-operator maps, including mappings to the Fock matrix from which multiple molecular observables can be simultaneously derived.

Auger spectra of thiouracils: Three molecules and four edges

This study presents a comprehensive investigation of the Auger spectra of three thionated uracils using a combination of theory and experiment. Non-resonant Auger spectra measured at the K-edges of carbon, nitrogen, and oxygen, as well as at the L2,3 edge of sulfur, are reported. Even though the final decay states (valence two-hole states) are the same for all edges, the resulting Auger spectra are strikingly different, highlighting the local nature of the Auger decay. Theoretical modeling, which employed a series of approaches incorporating varying levels of correlation and different treatments of the metastable states, provides a clear interpretation of the experimental spectra. The analysis highlights the crucial role of electron correlation, evident in both the density of the decay states and the spectra constructed using channel-specific partial decay widths, as well as the importance of spin–orbit coupling in shaping the L-edge spectra of thiouracils.