Designing novel ligands tailored to specific protein binding pockets remains a core objective in structure-based de novo drug design (SBDD). However, deep generative approaches encounter key challenges: standard graph neural networks fail to capture global pocket topology, SMILES-based models generate chemically invalid structures, and reinforcement learning remains unstable in multi-objective optimization. We evaluated our proposed framework on Janus Kinase 2 (JAK2) and Dopamine D2 Receptor (DRD2) targets. Compared to baseline models, our GPS-VAE successfully captured complex geometric dependencies, achieving robust active site representation and topological reconstruction. For molecular generation, we identified fragment-like and lead-like scaffolds demonstrating high predicted ligand efficiency (LE > 0.5) under the AutoDock Vina scoring function, discovering interesting macrocyclic adaptations targeting JAK2. We demonstrate that high-quality data representation combined with evolutionary search significantly enhances the efficiency of de novo drug design. We developed a new generative framework combining geometric deep learning and evolutionary search. First, we built a graph interaction transformer variational autoencoder (GPS-VAE) utilizing local graph attention networks and global transformer self-attention to extract physicochemical and geometric features. Second, we employed a Transformer-SELFIES autoencoder to replace the RNN-SMILES architecture, guaranteeing 100% chemical validity. Finally, we designed a variational projection network to anchor protein features into the chemical latent space, followed by structural refinement using the STONED evolutionary algorithm. Molecule preprocessing, docking, and fitness evaluations were performed using OpenBabel and AutoDock Vina.

An exploration into women's body image beliefs and sociocultural factors: A US/Italy cross-cultural comparison study.

Hyperleptinaemia is an emerging mechanism involved in the pathogenesis of obesity-related primary hypertension (HTN). This substudy from a large clinical trial aimed to determine differences in [leptin] in physically inactive adults (n = 69, 46.4% women) with HTN and living with overweight or obesity following a 16-week supervised aerobic exercise training (EX group, 2 days/week) intervention with an attention control (AC, only physical advice) group, both combined with a hypocaloric diet; to analyse whether the changes observed after the intervention were maintained after 6 months (6M); and to explore differences and potential mechanisms affected by sex. Cardiorespiratory fitness and body composition evaluation, and biochemical measurements were conducted. After the intervention [leptin] decreased in both groups (EX, 36%, p = 0.003; AC, 23%, p = 0.06), returning to baseline concentrations after 6M of follow-up in both groups. Changes over time were similar in both sexes, with no significant time × group interaction (p > 0.05). For all participants, body mass index was the only factor independently associated with [leptin] (β = 0.339, B = 2.11, SE = 0.53), explaining 12% of the variability (p < 0.001). Among women, waist-to-hip ratio (12%, β = 0.365, B = 135.8, SE = 59.4) and fat-free mass (11%, β = -0.443, B = -2.33, SE = -2.27) additionally and independently explained leptin variability (p < 0.05). A hypocaloric diet combined with supervised exercise appears to be an effective strategy for regulating [leptin] and improving metabolic health in individuals with HTN and overweight or obesity. The influence of fat distribution on leptin dynamics in women calls for research to consider sexual dimorphism.

Coverage-constrained multi-objective evolutionary recommendation algorithm for balancing accuracy, diversity, and novelty.

Introduction: Malingering, the intentional fabrication or exaggeration of symptoms for external gain, remains a significant challenge in forensic psychiatry. Psychiatric illnesses are particularly vulnerable to malingering due to their subjective symptom profiles and absence of objective biological markers. Early recognition is critical to protect genuine patients and maintain judicial integrity. Methods: We report three medico-legal cases referred by judicial authorities for psychiatric evaluation of fitness to stand trial. Each case was assessed through a multidisciplinary approach involving repeated clinical interviews, behavioral observation, review of medical records, collateral history, and standardized psychological assessments, including the Rorschach test and the Minnesota Multiphasic Personality Inventory (MMPI). Results: Indicators of malingering included initiation of treatment only after legal proceedings, inconsistencies between reported symptoms and observed behavior, and absence of expected side effects despite high-dose antipsychotics. Case I showed Ganser-like responses with intact Rorschach findings. Case II had elevated MMPI validity scales (F, Fp, FBS), consistent with over-reporting. Case III refused testing and displayed contradictions, such as claiming hallucinations without observable distraction. Together, these findings supported malingering driven by legal avoidance. Discussion: This series underscores the importance of multimodal evaluation in detecting malingering. Findings support The findings support the importance of multidisciplinary collaboration and the use of structured tools to improve diagnostic accuracy, prevent misuse of psychiatric diagnoses, and ensure justice in medico-legal contexts.

The article describes the teaching methods used by physical education teachers in the classroom setting and explores how observational learning, contextual learning, and flipped-classroom teaching methods interact in university physical education classes. It develops a multidimensional model of the application of various teaching methods in the field of university physical education. The MonoLoco method is utilized to attain the fuzzy localization process, and the information fusion method is utilized to improve human body posture data. Further, the HRNet model is introduced to include de-redundancy design and multi-resolution supervision to develop a keypoint detection algorithm for human skeletons based on the DHRNet network. By combining the monocular human-positioning algorithms with the DHRNet model to detect skeletal keypoints, the integrated physical fitness evaluation model for students is generated. The model is used to assess the performance of physical education teaching methods from a multidisciplinary perspective. The experimental results of integrating several teaching methods showed that the experimental group had improved their comprehensive evaluation index, including 50 meters, 1000 meters, standing long jumps, and shot put. The mean values of the 50 meters time reduced by 0.222 s, 1000 meters increased by 0.274 points, standing long jumps increased by 0.295 m, and shot put increased by 0.386 m. The above results indicate that the effectiveness of the physical education teaching with multiple teaching methods is considerable and applicable in university-level physical education.

The iterative search process of evolutionary algorithms (EAs) encapsulates optimization knowledge within historical populations and fitness evaluations. Effective utilization of this knowledge is crucial for facilitating knowledge transfer and online adaptation. However, current research typically addresses these goals in isolation and faces distinct limitations: evolutionary sequential transfer optimization often suffers from incomplete utilization of prior knowledge, while adaptive strategies, utilizing real-time knowledge, are limited to tailoring specific evolutionary operators. To simultaneously achieve these two capabilities, we introduce the Optimization Knowledge Adaptation Evolutionary Model (OKAEM), a unified learnable evolutionary framework capable of adaptively updating parameters based on available optimization knowledge. By parameterizing evolutionary operators via attention mechanisms, OKAEM enables learnable update rules that facilitate the utilization of optimization knowledge via two phases: pre-training to integrate extensive prior knowledge for efficient transfer, and adaptive optimization to dynamically update parameters based on real-time knowledge. Experimental results confirm that OKAEM significantly outperforms state-of-the-art sequential transfer methods across 12 transfer scenarios via pre-training, and surpasses advanced learnable EAs solely through its self-tuning mechanism in prior-free settings. Beyond demonstrating practical utility in prompt tuning for vision-language models, ablation studies validate the necessity of the learnable components, while visualization analyses reveal the model's capacity to autonomously discover interpretable evolutionary principles.

Structured physical education (PE) programs are essential for fostering students' physical development, cognitive performance, and emotional well-being in academic settings. This study introduces a novel intelligent decision-making algorithm (IDA) for the dynamic optimization and real-time assessment of college PE programs. The proposed framework synergistically integrates a hybrid genetic algorithm (GA) that is viable global optimizer with pattern search (PS) that is an adaptive local search technique. The exploration in the search domain is performed with GA while PS exploitation in a minimal computational budget is performed by the PS. The system encodes program performance data into chromosome-like representations, enabling nuanced evaluation across academic progress, health indices, and affective-psychomotor domains. A tailored vector of weight factors refines the fitness function to reflect individual learning trajectories and institutional goals. Experimental results demonstrate that the model achieves a high prediction accuracy of 98%, with quantifiable improvements of 151.13% in holistic performance, 19.63% in educational metrics, and 26.7% in psychomotor development as compared with reported results. Comparative models, including Random Forest (RF), Adaptive Neuro-Fuzzy Inference System (ANFIS), and RF regression, that achieved accuracies of 88.21%, 94.49%, and 82%, respectively. The hybrid framework maintained a mean global fitness value of 5.3451 × 10⁻¹² with an average computational time of 1357.04s over 100 runs that is used to validate the reliability of the proposed framework. By enhancing efficiency and reducing computational complexity, this AI-driven evaluation model offers a scalable and intelligent approach for real-time optimization and policy refinement in higher education PE curriculum planning.

Objective: To apply an individual and precise occlusal design method based on physiological tooth displacement for the digital design of posterior full crowns, and to explore its clinical effectiveness. Methods: This was a prospective randomized controlled study. From February 2024 to January 2025, patients who visited the Department of Prosthodontics, Peking University School and Hospital of Stomatology, with tooth defects in the second premolars or first molars and planned for full-crown restoration were enrolled. After tooth preparation, intraoral scanning was performed to obtain data of the maxillary and mandibular posterior dentitions and occlusal buccal surfaces on the side of the prepared tooth. Patients were randomly assigned using a random number method to undergo restoration design. Control: full crowns were designed according to a conventional protocol, using an automatically registered occlusal relationship based on the buccal surfaces. The occlusal clearance values were set based on previous studies and the clinical experience of the dental technician. Experimental: an individual occlusal design technique based on physiological tooth displacement was employed. Intraoral scan data underwent single-tooth segmentation and registration to derive virtual occlusal relationships in occlusion, followed by calculation of occlusal clearances of adjacent teeth. The occlusal clearance for the full crown was set as the mean value of the adjacent teeth's occlusal clearances. After fabrication of zirconia full crowns for both groups, they were delivered for clinical try-in. Occlusal evaluations were performed before and after occlusal adjustment. Primary outcomes included qualitative evaluation of occlusal fitness (qualitatively classified into three categories: acceptable occlusion, high occlusion and low occlusion) and occlusal adjustment height, while secondary outcomes comprised occlusal adjustment volume, articulating paper markings, occlusal clearance values, and occlusal adjustment time. Results: A total of 38 patients [12 males, 26 females; age: (39.5±13.2) years] were included in this study. Thirty-eight posterior full crowns were fabricated, with 19 allocated to the test group and 19 to the control group. The occlusal status before occlusal adjustment was as follows: in the test group, 4 crowns had acceptable occlusion, 15 had high occlusion, and none had low occlusion; in the control group, 1 crown had acceptable occlusion, 14 had high occlusion, and 4 had low occlusion. The number of cases with acceptable occlusion was significantly higher in the experimental group than in the control group (U=122.00, P=0.022). The occlusal adjustment height of the experimental group [(39.1±17.0) μm] was significantly lower than that of the control group [(79.5±50.2) μm] (t=-2.85, P=0.009). No significant differences were observed in adjustment time, volume, or articulating paper evaluation (all P>0.05). Conclusions: Using a virtual occlusal registration method for single-tooth segmentation based on the physiological tooth mobility and quantitatively referencing the occlusal contacts of adjacent teeth, this study significantly improved the occlusal suitability of full crowns and reduced the occlusal adjustment required.

Understanding disparities in physiological response to exercise load intensity in PE classes: Effects on physical fitness evaluation stratified by gender, grade and class type

Presentation of PASCA, a day hospital dedicated to multidisciplinary cancer survivorship assessment

Traditional approaches to managing changes in construction development projects are predominantly reactive, relying on manual progress tracking and heuristic decision-making, which often leads to significant cost overruns, schedule delays, and scope creep. This research addresses the critical need for a more predictive and data-driven management paradigm in the development of educational infrastructure. To overcome these limitations, this study develops and proposes a Method of proactive change management in development projects based on a Digital Twin (PCMM-DT). The research aim is to establish a closed-loop system where real-time site data, integrated via a Digital Twin (DT) and Building Information Modeling (BIM), allows for the dynamic simulation and mitigation of construction deviations before they disrupt project baselines. The core of the study is the development of a PCMM-DT method, which functions as an intelligent decision support system. The PCMM-DT method integrates a four-layer architecture – physical, digital twin, analytics, and action layers – to manage the project lifecycle. A PCMM-DT method algorithm was formulated to perform multi-objective optimization, balancing conflicting criteria such as lifecycle cost, functional utility, and project resilience. The evaluation is conducted using a PCMM-DT model for constraint handling, which validates proposed changes against geometric, safety, and quality standards. Furthermore, a PCMM-DT model for multi-objective fitness evaluation was constructed to quantify the impact of changes across economic, functional, and adaptability vectors. The application of the PCMM-DT method allows for the generation of a Pareto-optimal frontier, providing project managers with a robust, evidence-based ranking matrix for selecting the most appropriate mitigation strategies. This approach transforms the management of change orders from an adhoc process into a systematic, optimized workflow. The proposed PCMM-DT method represents a paradigm shift from reactive project governance to proactive, predictive management. By leveraging the PCMM-DT model for real-time analytics and decision support, project teams can identify deviations early, assess their impacts precisely, and select optimal configuration changes. This integration enhances the digital resilience of educational development projects, ensuring that capital investments remain aligned with both pedagogical objectives and budgetary constraints. Future research will focus on the empirical validation of the PCMM-DT method in large-scale infrastructure environments to further refine the predictive accuracy of the integrated digital twins.

Evolutionary reinforcement learning (ERL) offers a compelling alternative for continuous control by combining the population-level exploration of evolutionary algorithms with the gradient-based exploitation of reinforcement learning. However, applying conventional genetic operators to deep networks can be highly destructive, often inducing abrupt behavioral shifts that erase previously learned skills. Proximal distilled evolutionary reinforcement learning (PDERL) addresses this issue with phenotype-aware operators, leveraging proximal mutation and distillation crossover to produce safer and more constructive variations. Despite these advances, PDERL and many ERL frameworks still exhibit a fundamental evaluation asymmetry: an evolving actor population is guided by a single, centralized critic for fitness evaluation and action filtering. This single-critic dependence creates a bottleneck and a potential single point of failure, where bias or instability in value estimation can misdirect the evolutionary search. To overcome this limitation, we propose co-evolutionary proximal distilled evolutionary reinforcement learning (Co-PDERL), a heterogeneous dual-population framework that co-evolves both actor and critic populations. Co-PDERL extends phenotype-aware evolution to the value-function landscape via a loss-filtered distillation crossover and a Jacobian-based proximal mutation tailored for critics, and employs a condition-gated synchronization mechanism to enable robust bidirectional knowledge transfer between the evolutionary populations and the reinforcement learning agent. Experiments on MuJoCo continuous control benchmarks show that Co-PDERL outperforms competitive baselines on most tasks, including standard ERL and PDERL, improving both sample efficiency and asymptotic performance by effectively alleviating the single-critic bottleneck.

The fitness of the students in the college is a critical parameter of their health, well-being, and academic output. Nonetheless, the assessment of physical fitness determinants is uncertain and subjective due to differing opinions among experts and the interconnection of the standards with one another. In addressing this complexity, the current research presents a new Fermatean Fuzzy (FF) Deck-of-Cards-Method (DoCM) Multi-Criteria Decision-Making (MCDM) model for the quantitative evaluation of physical fitness in academic settings. The model is based on the fact that FF sets have high expressiveness, enabling them to represent expert hesitation and dual uncertainty more effectively than conventional fuzzy models. The DoC approach is used to obtain the systematic and preference-based weights of the evaluation criteria, thereby achieving rational and flexible prioritization. A hypothetical case study is created to demonstrate the relevance of the framework, encompassing criteria such as endurance, strength, flexibility, body composition, and mental well-being. The findings underscore the model’s ability to produce consistent and explainable rankings of fitness determinants using incomplete or unclear information. When compared to current literature where most models use FS/IFS/PFS as their default uncertainty modeling framework with cognitively demanding weighting functions, the proposed FF-DoC framework bridges the gap of integrating high-expressiveness uncertainty modeling with transparent and human-eliciting weight elicitation to assess institutional fitness competency.

Introduction: There are divergences about the effects of obesity on plasma concentrations of interleukin-7 (IL-7) and the role of physical exercise as a metabolic regulator of this cytokine. Objective: To compare the plasma concentrations of IL-7 between obese and normal weight individuals, in addition to evaluating the effects of combined training (CT) on concentrations of this cytokine in obese individuals. Methods: Initially, obese group (OG, n = 15) and normal weight group (NWG, n = 8) were compared. Subsequently, the obese individuals participated in a pre-post 16-week experimental period and were randomly distributed into obese CT group (OT, n = 8) and obese control group (OC, n = 7). Physical fitness, body composition, and IL-7 concentrations evaluations were performed. The CT program consisted of strength training and aerobic training in the same session. Results: A significant increase of IL-7 was observed in the OG (27.14 ± 3.64 pg/mL) compared to the NWG (21.26 ± 3.93 pg/mL) (p = 0.01). Although no significant group x time interaction was found, a time effect was observed in the pre-post experimental period on IL-7 concentrations (OC: 18.8%, ES: 1.68; OT: 28.5%, ES: 5.15). In the assessment of the effect size, a greater reduction in OT was observed. Moreover, reductions in weight, body mass index and fat mass were observed in the OT group when compared to CG, accompanied by significant increases in 1 maximum repetition test in the Leg Press and Bench Press, and maximum oxygen consumption. Conclusion: Obese individuals have higher circulating concentrations of IL-7, which suggest that elevated body weight and fat are associated with an increase in this cytokine. Although the effect size for IL-7 was larger in the OT group, probabilistic statistics did not show a significant effect.

ABSTRACT Tuberculosis (TB), caused by Mycobacterium tuberculosis ( M. tb ), is a primary contributor to morbidity and mortality among underdeveloped countries. The spread of drug‐resistant TB is one crucial concern to world health. Mutations in rpoB, the beta subunit of M. tb 's DNA‐directed RNA polymerase (PDB ID: 6DVC), have been identified as a critical source of rifampicin (RIF) resistance. To find efficient small molecule compounds from Natural Compound Archives, this work models resistant rpoB mutants (S450L and D435V) and uses pharmacophore‐based virtual screening to pick out molecules of interest. Phase fitness evaluations from the pharmacophore‐based virtual screening were used and created 84 hits from the three‐dimensional hypothesis screening. The docking results revealed that asn:98397 more effectively inhibited the wild‐type rpoB with a score of −9.8 kcal/mol, and also the mutations S450L and D435V had interactions with scores of −9.295 and −9.670 kcal/mol, respectively. On the contrary, RIF had less interaction with the wild type: −2.767 kcal/mol, S450L: −6.676 kcal/mol, and D435V: −8.241 kcal/mol, revealing the incapability of RIF inhibition and thus showing resistance. Molecular dynamics (MD) simulations were performed on leading hit complex structures to investigate their rigidity, interconnections, and longevity. The identified phytocompound can behave as an alternative to RIF and overcome resistance.

This paper introduces the Self-Adaptive Cuckoo Genetic Algorithm (SACGA) for Echo State Network (ESN) optimization. Traditional Genetic Algorithms (GAs) often simplify the search space or focus solely on error minimization, overlooking key factors such as the reservoir quality of ESN. The proposed SACGA addresses these limitations by optimizing all hyperparameters across the entire search space and incorporating the concept of separation, as a reservoir quality metric, in the fitness evaluation alongside the standard predictive error minimization criterion. The SACGA features a novel design that integrates a self-adaptive mutation mechanism and Cuckoo-inspired offsprings within a traditional GA framework, for an optimal balance between exploration and exploitation. The self-adaptive mechanism dynamically adjusts the mutation rates, affecting both the crossover and mutation processes. The Cuckoo-inspired offsprings, generated via the Lévy flight mechanism, explore the search space’s boundaries. The SACGA outperforms baseline GAs (with fixed mutation rates of 10% and 1%, and a deterministic mutation rate scheme) in Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) metrics on benchmark datasets (Multiple Superimposed Oscillator, Mackey-Glass, Nonlinear Auto-Regressive Moving Average). Results show faster convergence to lower error rates highlighting its efficiency in ESN optimization for time series prediction.

Noncoherent sampling in digital storage oscilloscopes introduces spectral leakage and picket-fence effects, which bias harmonic parameter estimates and can obscure weak components in the vicinity of strong tones. This work presents a multi-stage frequency-correction framework for harmonic and interharmonic analysis, consisting of candidate-interval screening in the spectral domain, continuous frequency refinement with a fitness-guided search, least-squares amplitude/phase estimation, and an iterative reconstruct-subtract procedure to progressively suppress leakage-induced interference. A central processing unit-graphics processing unit heterogeneous implementation is further developed to exploit fine-grained data parallelism in interval extraction and batched fitness evaluations. Experimental results demonstrate that the proposed method achieves markedly higher frequency accuracy than a windowed fast Fourier transform baseline under noncoherent conditions (mHz-level maximum absolute error, exceeding 37× improvement over the baseline). Across multi-harmonic cases, it attains accuracy comparable to multiple signal classification (MUSIC) while avoiding the need for sensitive subspace-order selection and other strong prior choices, thereby improving deployment robustness. In challenging adjacent-tone and large-dynamic-range scenarios, where interpolated discrete Fourier transform and MUSIC may exhibit pronounced sensitivity to residual leakage, the proposed iterative cancellation substantially stabilizes weak-component estimation (e.g., amplitude error reduced from tens of percent to the 0.2% level). The framework also supports a continuous accuracy-cost trade-off via the search-budget parameters, providing a practical path toward reliable oscilloscope-resident harmonic/interharmonic measurements.

The article is devoted to the analysis of European Union countries’ experience in using physical fitness tests within the school education system, with particular attention to their role in monitoring children’s health status and motor development. The relevance of the study is determined by the growing importance of evidence-based approaches to physical education assessment and the need to modernize national systems of physical fitness evaluation in line with contemporary European practices. The aim of the article is to identify and systematize practical approaches to the organization, content, and interpretation of physical fitness testing results in schools of EU countries. The study is based on the analysis and synthesis of scientific publications, official reports of the European Commission, materials of international initiatives and projects (EUROFIT, ALPHA Fitness Test, FitBack), as well as descriptions of national and regional testing programs implemented in Germany, Austria, and Poland. The results indicate that modern European models of physical fitness testing are predominantly oriented toward a health-related paradigm rather than normative control. These models are characterized by a high level of procedural standardization, the use of validated and reliable test batteries, age- and sex-specific reference values, and percentile-based interpretation of results. A significant trend identified is the active digitalization of data collection, processing, and feedback, which enhances the comparability of results at national and supranational levels and supports long-term monitoring. The generalized European experience demonstrates that systematic physical fitness testing can be effectively integrated into the educational process without excessive burden on students and teachers, provided that appropriate organizational conditions, specialist training, and clear communication with stakeholders are ensured. The findings may serve as a methodological basis for improving national systems of physical fitness assessment and for adapting European monitoring approaches within school physical education.

Autonomous vehicle development demands reliable control strategies that can operate under dynamic and safety-critical conditions. Simulation-based evaluation plays a vital role by enabling controlled, repeatable testing without the risks of real-world deployment. This paper presents a custom-built self-driving car simulation framework that integrates feedforward neural networks with genetic algorithms for autonomous control optimization. Implemented entirely in JavaScript without external dependencies, the framework evolves steering and acceleration policies through neuro-evolution using sensor-based perception. A lightweight sensor configuration and multi-objective fitness evaluation enable smooth navigation and effective collision avoidance across complex track layouts. Experimental results demonstrate a navigation success rate of 91.7 percent and a collision avoidance accuracy of 96.8 percent after 47 generations, while achieving faster convergence than gradient-based training methods. The proposed framework offers a flexible and computationally efficient platform for autonomous vehicle research and rapid prototyping.