Category: Nevin Manimala

Sci Rep. 2025 Aug 11;15(1):29376. doi: 10.1038/s41598-025-15057-x.

ABSTRACT

This study investigates emergent behaviors in multi-agent pursuit-evasion games within a bounded 2D grid world, where both pursuers and evaders employ multi-agent reinforcement learning (MARL) algorithms to develop adaptive strategies. We define six fundamental pursuit actions-flank, engage, ambush, drive, chase, and intercept-which combine to form 21 types of composite actions during two-pursuer coordination. After training with MARL algorithms, pursuers achieved a 99.9% success rate in 1,000 randomized pursuit-evasion trials, demonstrating the effectiveness of the learned strategies. To systematically identify and measure emergent behaviors, we propose a K-means-based clustering methodology that analyzes the trajectory evolution of both pursuers and evaders. By treating the full set of game trajectories as statistical samples, this approach enables the detection of distinct behavioral patterns and cooperative strategies. Through analysis, we uncover emergent behaviors such as lazy pursuit, where one pursuer minimizes effort while complementing the other’s actions, and serpentine movement, characterized by alternating drive and intercept actions. We identify four key cooperative pursuit strategies, statistically analyzing their occurrence frequency and corresponding trajectory characteristics: serpentine corner encirclement, stepwise corner approach, same-side edge confinement, and pincer flank attack. These findings provide significant insights into the mechanisms of behavioral emergence and the optimization of cooperative strategies in multi-agent games.

PMID:40789927 | DOI:10.1038/s41598-025-15057-x

Nat Genet. 2025 Aug 11. doi: 10.1038/s41588-025-02286-z. Online ahead of print.

ABSTRACT

Mixed-model association analysis (MMAA) is the preferred tool for performing genome-wide association studies. However, existing MMAA tools often have long runtimes and high memory requirements. Here we present LDAK-KVIK, an MMAA tool for analysis of quantitative and binary phenotypes. LDAK-KVIK is computationally efficient, requiring less than 10 CPU hours and 5 Gb memory to analyze genome-wide data for 350,000 individuals. Using simulated phenotypes, we show that LDAK-KVIK produces well-calibrated test statistics for both homogeneous and heterogeneous datasets. When applied to real phenotypes, LDAK-KVIK has the highest power among all tools considered. For example, across 40 quantitative UK Biobank phenotypes (average sample size 349,000), LDAK-KVIK finds 16% more independent, genome-wide significant loci than classical linear regression, whereas BOLT-LMM and REGENIE find 15% and 11% more, respectively. LDAK-KVIK can also be used to perform gene-based tests; across the 40 quantitative UK Biobank phenotypes, LDAK-KVIK finds 18% more significant genes than the leading existing tool. Last, LDAK-KVIK produces state-of-the-art polygenic scores.

PMID:40789918 | DOI:10.1038/s41588-025-02286-z

Npj Ment Health Res. 2025 Aug 11;4(1):37. doi: 10.1038/s44184-025-00150-w.

ABSTRACT

The aim was to assess the effectiveness of an intervention comprising a teacher-training program and the implementation of a ten-session mindfulness-based intervention (MBI) in regular classroom teaching for 16- to 24-year-old students in Denmark. In a cluster-randomized trial (2019-2021), upper secondary schools and vocational schools for social and health care were randomly assigned 1:1 to receive the intervention (21 schools, 35 teachers, 438 students) or teaching-as-usual (22 schools, 38 teachers, 551 students). Eighteen self-report measures of mental health were collected at baseline, three, and six months, with the Warwick-Edinburgh Mental Well-Being Scale (SWEMWBS) as the primary outcome. Intention-to-treat analyses were conducted using mixed-effects linear regression and bootstrapping. Fourteen intervention schools (n = 277) and 17 teaching-as-usual schools (n = 419) provided follow-up data. No statistically significant effect on the primary outcome (SWEMWBS) was found in the total population. However, a small positive effect on SWEMWBS was observed among females in upper secondary schools at three months, but not sustained at six months. Trial Registration: ClinicalTrials.gov Identifier: NCT04610333, registered October 26, 2020, https://clinicaltrials.gov/ct2/results?cond=&term=NCT04610333&cntry=&state=&city=&dist= .

PMID:40789917 | DOI:10.1038/s44184-025-00150-w

Sci Rep. 2025 Aug 11;15(1):29439. doi: 10.1038/s41598-025-14784-5.

ABSTRACT

With the rapid growth in the exchange of digital data, the problem of protecting images has become essential, particularly in sectors such as medicine, surveillance and secure communications. Traditional encryption techniques, such as DES, AES and RSA, are effective for text, but often ineffective for images, due to high redundancy and high correlation between pixels. We suggest a novel image encryption algorithm based on chaotic maps and the variable-step Josephus problem to get over these restrictions and increase the encryption resilience. This algorithm uses Kepler Chaotic Optimisation Algorithm (CKOA) to select the most suitable chaotic maps, guaranteeing optimal diffusion and complex pixel confusion. Key generation is enhanced with MD5 and SHA-256 hash functions, providing increased resistance to collisions and attacks. In addition, Discrete Wavelet Transform (DWT) is integrated to compress images and reduce processing time, while maintaining a high average entropy of 7.999 for encrypted images. Experimental tests demonstrate strong resistance against cryptographic attacks, including 99.6% as pixel change rate and 33.31% as unified average change intensity, ensuring optimum security. Compared with other image-based encryption schemes, our method stands out for its speed and reduced computational complexity, while offering superior security.

PMID:40789914 | DOI:10.1038/s41598-025-14784-5

Sci Rep. 2025 Aug 11;15(1):29420. doi: 10.1038/s41598-025-15388-9.

ABSTRACT

Early prediction of academic performance is vital for reducing attrition in online higher education. However, existing models often lack comprehensive data integration and comparison with state-of-the-art techniques. This study, which involved 2,225 engineering students at a public university in Ecuador, addressed these gaps. The objective was to develop a robust predictive framework by integrating Moodle interactions, academic history, and demographic data using SMOTE for class balancing. The methodology involved a comparative evaluation of seven base learners, including traditional algorithms, Random Forest, and gradient boosting ensembles (XGBoost, LightGBM), and a final stacking model, all validated using a 5-fold stratified cross-validation. While the LightGBM model emerged as the best-performing base model (Area Under the Curve (AUC) = 0.953, F1 = 0.950), the stacking ensemble (AUC = 0.835) did not offer a significant performance improvement and showed considerable instability. SHAP analysis confirmed that early grades were the most influential predictors across top models. The final model demonstrated strong fairness across gender, ethnicity, and socioeconomic status (consistency = 0.907). These findings enable institutions to identify at-risk students using state-of-the-art interpretable and fair models. These findings enable institutions to identify at-risk students using state-of-the-art, interpretable, and fair models, advancing learning analytics by validating key success predictors against contemporary benchmarks.

PMID:40789907 | DOI:10.1038/s41598-025-15388-9

Clin Nutr. 2025 Aug 5;52:215-224. doi: 10.1016/j.clnu.2025.07.034. Online ahead of print.

ABSTRACT

BACKGROUND & AIMS: Anorexia nervosa (AN) is a severe eating disorder with significant somatic complications, potentially involving an altered intestinal barrier. Prior studies yielded heterogeneous results, ranging from decreased to increased intestinal permeability in AN. This study aimed to assess intestinal permeability in adolescent patients with AN by comparing serum markers Zonulin, Lipopolysaccharide-binding protein (LBP), and intestinal Fatty acid-binding protein (iFABP) with healthy controls (HC) over time.

METHODS: We conducted a longitudinal study on 60 female adolescent patients with AN and 36 age-matched HC. Blood samples were collected at admission to inpatient treatment, at discharge, and at a one-year follow-up. Zonulin, LBP, iFABP and inflammation marker Interleukin 6 levels were measured alongside clinical parameters. Linear mixed models were applied to assess group differences and longitudinal changes. Pearson correlations were used to explore clinical associations.

RESULTS: Zonulin levels were significantly decreased in the AN group at admission and again at the follow-up compared to HC. LBP levels were consistently reduced in AN across all three time points. These alterations thus persisted despite weight recovery. No significant differences were observed in iFABP levels between the groups or over time. Zonulin and LBP positively correlated with Interleukin 6, linking gut permeability to inflammatory processes.

CONCLUSION: Decreased Zonulin and LBP levels suggest reduced paracellular intestinal permeability in adolescent AN while transcellular permeability (iFABP) appears to be rather unaffected. The persistence of these alterations despite weight recovery suggests that reduced intestinal permeability may be a trait rather than a state marker in AN. Our findings challenge the concept of a “leaky gut” in adolescent AN.This highlights the need for further research on the intestinal barrier in AN, while also integrating the role of nutritional intake, gut microbiome interaction and immune reactions, to support the development of gut-targeted therapies.

PMID:40789230 | DOI:10.1016/j.clnu.2025.07.034

Poult Sci. 2025 Jul 23;104(10):105596. doi: 10.1016/j.psj.2025.105596. Online ahead of print.

ABSTRACT

Poultry farming in Gujarat faces persistent challenges from infectious diseases, foodborne pathogens, and the growing threat of antimicrobial resistance (AMR), necessitating the adoption of stringent biosecurity measures encompassing isolation, traffic control, and sanitation. The study aimed to evaluate the knowledge and effectiveness of biosecurity practices among line supervisors involved in contractual commercial broiler farming in Gujarat. This study evaluated the socio-economic characteristics and biosecurity knowledge of 33 line supervisors through a one-day capacity-building training program. Pre- and post-training assessments, including structured questionnaires and Likert scales, were analyzed using descriptive statistics, and Factor Analysis of Mixed Data (FAMD) in R (v4.0.2). Results revealed that most participants were young (mean age 33.94), married males with higher secondary or graduate education and over a decade of poultry farming experience. Post-training evaluations showed a significant improvement in biosecurity knowledge and practices, with 75.76 % rating the training as excellent and was recognized as useful by 90.91 % of respondents. Furthermore, for the non-implementation of comprehensive biosecurity measures by the farmers, lack of awareness was the major attribute given by line supervisors (57.58 %) for the farmers, followed by inadequate control (51.52 %) and insufficient knowledge (48.48 %). Additionally, FAMD identified 76 epidemiological points across breeding farms, commercial broiler farms, and markets, offering insights into zoonotic risk and poultry distribution network (PDN) structure. The findings investigated the need for structured training and advisory support, along with government-led financial incentives, to strengthen biosecurity implementation in poultry systems.

PMID:40789226 | DOI:10.1016/j.psj.2025.105596

J Proteome Res. 2025 Aug 11. doi: 10.1021/acs.jproteome.4c01114. Online ahead of print.

ABSTRACT

Motivated by the lack of adequate tools to perform pathway enrichment analysis, this work presents an approach specifically tailored to SomaScan data. Starting from annotated gene sets, we developed a greedy, top-down procedure to iteratively identify strongly intracorrelated SOMAmer modules, termed “SomaModules”, based on 11K SomaScan data. We generated two repositories based on the latest MSigDB and MitoCarta releases, containing more than 40,000 SOMAmer-based gene sets combined. These repositories can be utilized by any unstructured pathway enrichment analysis tool. We validated our results with two case examples: (i) Alzheimer’s disease specific pathways in a 7K SomaScan case-control study, and (ii) mitochondrial pathways using 11K SomaScan data linked to physical performance outcomes. Using gene set enrichment analysis (GSEA), we found that, in both examples, SomaModules had significantly higher enrichment than the original gene set counterparts. These findings were robust and not significantly affected by the choice of enrichment metric or the Kolmogorov enrichment statistic used in the GSEA procedure. We provide users with access to all code, documentation and data needed to reproduce our current repositories, which also will enable them to leverage our framework to analyze SomaModules derived from other sources, including custom, user-generated gene sets.

PMID:40789203 | DOI:10.1021/acs.jproteome.4c01114

Psychiatriki. 2025 Aug 5. doi: 10.22365/jpsych.2025.022. Online ahead of print.

ABSTRACT

The present study evaluated the impact of a hybrid Assertive Community Treatment (ACT) model of care on the direct medical costs of treating severe mental illness (SMI) patients in rural Greece. The study aimed to determine whether this model resulted in significant cost differences compared to usual treatment while also assessing its cost-effectiveness based on clinical improvements. A total of 23 patients with SMI and multiple hospitalizations were followed up for 16 months under the hybrid ACT model. Direct medical costs were estimated using previously published Greek data on schizophrenia treatment costs. Cost differences before and after the implementation of the hybrid ACT model were calculated, and cost-effectiveness was assessed using the Incremental Cost-Effectiveness Ratio (ICER), which reflects the cost per unit increase in Global Assessment of Functioning (GAF) scores. There was no statistically significant difference in direct medical costs between usual care (310,029€) and hybrid ACT care (313,896€), with a small cost increase of 3,867€ (p = 0.077). However, hybrid ACT care significantly reduced hospitalizations and length of inpatient stay, leading to an 86.9% reduction in total inpatient days. Clinical improvements were also observed, with GAF scores increasing from 40.43 to 47.26. Cost-effectiveness analysis demonstrated a particularly low ICER of 25.9€ per GAF point gained, suggesting a cost-efficient intervention. In an alternative scenario, the 2024 pricing was estimated with the use of the Consumer Price Index. In this case, the hybrid ACT care appeared to be significantly cost-saving by 25.5%. A rough estimation of indirect costs revealed further cost savings in favor of the hybrid ACT. The hybrid ACT model proved to be cost-effective due to its strong impact on reducing inpatient care and improving patient functioning. These findings align with international studies demonstrating the economic and clinical benefits of community-based mental health care. Future research should focus on larger, multicenter studies to confirm cost-effectiveness and explore the impact on indirect costs, such as caregiver burden and law enforcement involvement. The results support further investment in hybrid ACT services in rural Greece to enhance mental health care delivery in low-resourced settings.

PMID:40789197 | DOI:10.22365/jpsych.2025.022

Anal Chem. 2025 Aug 11. doi: 10.1021/acs.analchem.5c03797. Online ahead of print.

ABSTRACT

Pulsed nanoelectrospray ionization (nanoESI) mass spectrometry has been proven to have several merits compared to its direct current (DC) counterparts, such as high sensitivity, high matrix tolerance, and low electrochemical interference. However, the reliance on external devices like signal generators and power amplifiers limits the technology’s suitability for on-site use. Here, we developed a frequency-amplitude tunable hand-held pulsed high-voltage supply specifically designed for nanoESI. This portable detection system eliminates reliance on bulky ancillary equipment of conventional approaches, with a total mass of 185 g and an ergonomic hand-held configuration. Through systematic modulation of frequency (10-50 kHz) and voltage (1.5-6 kV), the device achieves optimal detection performance across diverse analytical standards and complex matrixes, including serum and cellular samples. The pulse-induced self-cleaning effect prevented nozzle clogging while maintaining functionality under high-salt conditions (855 mM NaCl). To demonstrate clinical applicability, we analyzed 198 serum specimens (100 from atherosclerotic cardiovascular disease patients (ASCVD) and 98 age-matched controls), identifying statistically significant dysregulation of multiple metabolites in the ASCVD cohort and achieving a diagnostic accuracy of 85%. Field deployment trials using portable mass spectrometers further validated the system’s potential for rapid on-site diagnostics, establishing a new paradigm for point-of-care analytical technology.

PMID:40789175 | DOI:10.1021/acs.analchem.5c03797