1 IntroductionCerebral infarction represents a relatively uncommon yet prognostically significant complication in the perioperative setting, with a documented incidence of 0.9% following thoracic surgical procedures [1]. While conventional etiologies such as atherosclerotic thromboembolism and cardioembolic events secondary to atrial fibrillation account for most postoperative cerebral infarctions, pulmonary surgery introduces a distinct mechanism: pulmonary vein stump thrombosis (PVST). This condition arises when thrombotic material formed at the surgically resected pulmonary vein stump migrates retrograde into the left atrium, subsequently embolizing to the cerebral artery [2].Despite its catastrophic neurological consequences, PVST-related AIS remains underrecognized in clinical practice, often leading to delayed diagnosis and suboptimal management. This case report illustrates a characteristic presentation of PVST-induced AIS, accompanied by a systematic review of literature regarding its pathogenesis and evidence-based management strategies, aiming to enhance clinical vigilance and therapeutic decision-making.

Background: Glycosyltransferases (GTs) are principal post-reactive modifying enzymes responsible for establishing natural glycosidic bonds in secondary metabolites, playing a critical regulatory role in plant cellular metabolic homeostasis. Biphenyl, dibenzofuran, and their glycosides, the most abundant phytoalexins in the apple subfamily, are synthesized de novo after infection by bacteria or fungi. Nevertheless, the biological functions of GTs in Sorbus aucuparia remain largely uncharacterised. Purpose: This study systematically evaluated the impact of Sorbus aucuparia uridine diphosphate glycosyltransferases (SaUGTs) on biphenyl phytoalexin metabolism and growth patterns in Sorbus aucuparia suspension cells (SASCs) under yeast extract (YE)-induced biotic stress. Methods: The study established standardized SASCs cultures with controlled induction protocols using YE for biotic stress simulation. A multi-omics framework integrated phenotypic analyses, targeted metabolomics (UPLC-QTOF-MS), transcriptional profiling (quantitative PCR), and enzymatic functional assays. Results: YE treatment induced a biomass decline in SASCs, coinciding with substantial accumulation of biphenyl derivatives and glycosides. Temporal profiling revealed dynamic fluctuations in metabolite concentrations, reflecting sequential biosynthetic transformations. Stress exposure elevated soluble protein content and significantly up-regulated SaUGTs expression. YE-induced SaUGTs promote glycosylation of de novo-synthesised biphenyl phytoalexins (noraucuparin, aucuparin) and 2′-hydroxyaucuparin, with optimal cell growth occurring during metabolic equilibrium between aglycones and glycosides. Conclusion: These findings suggest a previously unrecognised regulatory strategy, whereby SASCs alleviate biotic stress through GT-mediated maintenance of phytoalexin-glycoside homeostasis, thus preventing detrimental over-activation of defence mechanisms.

Chikitsak: An Integrated Healthcare Assistant for Symptom Analysis and Radiological DiagnosisAbstractThis paper introduces Chikitsak, an integrated healthcare assistant that combines a symptom-based conversational interface with radiological diagnostic modules. The system collects basic user information and, via an interactive chatbot, gathers detailed symptom data. Based on user input, the chatbot provides three probable disease predictions along with precautionary measures, dietary recommendations, and suggested medications. In parallel, the system offers automated diagnostic support for pneumonia and brain tumor detection by analyzing chest X-ray and brain MRI images using deep learning models. Experimental evaluation shows promising accuracy, and the integration of natural language processing (NLP) with radiological image analysis can significantly aid early disease detection and patient engagement.KeywordsHealthcare, Chatbot, Natural Language Processing, Deep Learning, Pneumonia Detection, Brain Tumor Detection, Medical Imaging, AI in Healthcare1. IntroductionThe rapid growth of artificial intelligence (AI) and deep learning in healthcare has catalyzed the development of systems that enhance early disease detection and patient management. Chikitsak leverages these technologies by integrating a symptom-based chatbot with state-of-the-art diagnostic models. The chatbot collects patient demographics and symptom descriptions, engages in a structured query process, and provides preliminary diagnoses with actionable recommendations. Complementing this, specialized models assess radiological images to detect pneumonia and brain tumors, offering a dual approach to patient screening. Such systems hold the potential to reduce the burden on healthcare professionals and improve patient outcomes by providing timely and accessible preliminary assessments.2. Related Work2.1 Chatbots in HealthcareRecent research demonstrates that chatbots can effectively deliver preliminary health advice, triage patients, and improve healthcare accessibility. Studies such as “Chatbots in Healthcare: A Systematic Review” have underscored the benefits of conversational agents in patient engagement and symptom assessment (NCBI PMC). Chatbots have been shown to reduce the workload on clinical staff while offering consistent, round-the-clock support.2.2 Deep Learning for Radiological DiagnosisDeep learning models have revolutionized medical imaging analysis. For instance, CheXNet, a convolutional neural network, achieved radiologist-level performance in detecting pneumonia from chest X-rays (CheXNet, arXiv). Similarly, CNN-based approaches have been applied successfully to brain MRI data for tumor detection, achieving high sensitivity and specificity (NCBI PMC). These advances provide a solid foundation for integrating radiological diagnostics into a comprehensive patient assessment system like Chikitsak.3. Methodology3.1 System ArchitectureThe Chikitsak system is divided into two primary modules: Symptom Analysis Module (Chatbot): User Data Collection: Users provide basic information (name, age, sex) through an initial form. Conversational Interface: A chatbot engages the user in a dynamic dialogue, asking a series of five targeted questions based on initial symptom input. Natural language processing (NLP) techniques are applied to understand and classify symptoms. Preliminary Diagnosis: After gathering sufficient data, the system predicts three potential diseases, accompanied by recommended precautions, dietary suggestions, and medications. A downloadable report is generated for user reference and clinical consultation. Radiological Diagnosis Module: Image Upload and Preprocessing: Users can upload chest X-ray and brain MRI images. Model Inference: Two separate convolutional neural network (CNN) models are deployed: Pneumonia Detection: Trained on large-scale chest X-ray datasets, the model analyzes X-ray images to predict pneumonia. Brain Tumor Detection: A dedicated CNN model processes MRI images to detect brain tumors. Output Generation: The results, including probability scores and visualizations of the detection areas, are incorporated into the final downloadable diagnostic report. 3.2 Data and Model Training Chatbot Training: The NLP engine is trained on medical conversation datasets to improve its ability to recognize symptom descriptions and contextual nuances. Data augmentation techniques are applied to ensure robust performance in varied linguistic expressions. Radiological Models: Pneumonia Model: Leveraging architectures similar to CheXNet, the model is fine-tuned using publicly available chest X-ray datasets (e.g., ChestX-ray14). Brain Tumor Model: The MRI-based model is developed using annotated brain imaging datasets available from sources such as Kaggle or institutional repositories. Both models undergo rigorous cross-validation and testing to ensure accuracy, sensitivity, and specificity.3.3 Integration and User WorkflowThe front-end interface guides users sequentially through the system: Initial Data Entry: User demographics and symptom input. Chatbot Interaction: The conversational agent asks follow-up questions to refine symptom details. Diagnostic Prediction: Based on the dialogue, the system predicts possible diseases. Image Analysis (Optional): For users opting for further diagnostics, the radiological module processes uploaded images. Report Generation: A consolidated report detailing the preliminary diagnoses, recommended next steps, and radiological analysis (if applicable) is generated and made available for download. The seamless integration of these modules ensures a comprehensive preliminary assessment that can be shared with healthcare professionals.4. Experimental Evaluation4.1 Evaluation MetricsFor the chatbot module, performance is measured using: Precision and Recall: Accuracy in predicting disease categories. User Satisfaction Scores: Evaluated through usability studies and feedback. For the radiological modules, the following metrics are used: Accuracy, Sensitivity, and Specificity: Standard performance metrics for medical image classification. Area Under the Receiver Operating Characteristic Curve (AUC-ROC): To evaluate model discrimination capabilities. 4.2 ResultsPreliminary experiments indicate that the chatbot achieves high accuracy in symptom classification, with user satisfaction ratings suggesting ease of use and clarity in communication. The pneumonia detection model demonstrates performance comparable to established benchmarks such as CheXNet (CheXNet, arXiv), while the brain tumor detection model shows promising accuracy, with further improvements anticipated through additional training on more diverse datasets.4.3 DiscussionThe integration of a symptom-based chatbot with radiological diagnostics provides a multi-modal approach to early disease detection. Although the system is not intended to replace professional medical advice, it serves as an effective triage tool, guiding users towards timely clinical evaluation. Challenges remain in ensuring robust performance across diverse populations and addressing data privacy concerns. Future iterations will focus on expanding the disease prediction library and refining image analysis algorithms.5. ConclusionChikitsak represents a significant step towards the integration of AI-driven tools in healthcare. By combining NLP-powered symptom analysis with deep learning-based radiological diagnostics, the system offers a comprehensive, user-friendly interface for preliminary disease assessment. While initial results are promising, ongoing work will refine the models, extend functionality, and further validate the system through clinical trials. This approach not only has the potential to enhance early diagnosis but also to empower patients with actionable health information.6. Future WorkFuture enhancements for Chikitsak include: Expansion of Diagnostic Capabilities: Incorporating additional disease models beyond pneumonia and brain tumors. Clinical Trials: Conducting rigorous clinical evaluations to validate the system's efficacy and safety. User Experience Enhancements: Improving the chatbot’s conversational depth and integrating multi-language support. Data Security: Strengthening data privacy measures and compliance with healthcare regulations (e.g., HIPAA, GDPR). References CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays with Deep Learning. Retrieved from https://arxiv.org/abs/1711.05225 Brain Tumor Detection Using Convolutional Neural Network. NCBI PMC Article Chatbots in Healthcare: A Systematic Review. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6460224/ WHO Guidelines on Digital Health Interventions. Retrieved from https://www.who.int/publications/i/item/9789241511642

Information regarding the age of amphibians is essential for species management, particularly with respect to invasive alien species. American bullfrog (Aquarana [Lithobates] catesbeianus) is a representative invasive alien species that are affecting ecosystems, but information on their age and growth rates is lacking. This study aimed to estimate the age and growth curve of A. catesbeianus in South Korea. A total of 49 individuals of varying sizes and sexes were collected from 10 different sites between 2019 and 2023. All individuals measured the snout-vent length (SVL) and estimated their age via skeletochronology. The analysis revealed distinct lines of arrested growth (LAGs) in bone cross-sections, allowing for accurate age estimation. The age of males was estimated to be 1- 5 years, females 0 - 4 years, and juveniles 0 - 2 years. No significant differences in age and SVL were found between sexes; however, a significant difference was observed between adults and juveniles. Given the limited sample size, the growth curve was estimated using only male individuals, revealing rapid growth up to two years of age, followed by a deceleration in growth rate. The asymptotic size and growth rate of A. catesbeianus was significantly larger than that of native amphibians, it could pose a substantial threat to native amphibians. These findings highlight the need for early monitoring and population control measures to mitigate the ecological impact of this invasive species.

This short communication develops a variational formulation for modeling a compressible electronic fluid motion. The results are based on standard tools of calculus of variations and optimization theory. The context here addressed is essentially an Euler-Bernoullian one and includes also a new approximate Bernoulli-perfect gas equation.

The rapid proliferation of quadcopters in various applications, from surveillance to delivery services, has necessitated the development of efficient fault detection systems to ensure operational safety and reliability. This study explores the use of Artificial Intelligence (AI) and acoustic data for automated fault detection in quadcopter propellers. By leveraging machine learning algorithms, the proposed system analyzes acoustic signatures generated by propellers during operation to identify anomalies such as cracks, imbalances, or deformations. A dataset of acoustic signals was collected from both healthy and faulty propellers under varying operational conditions. Feature extraction techniques, including spectral analysis and time-frequency transformations, were employed to preprocess the data. The processed data was then used to train and validate AI models, including Convolutional Neural Networks (CNNs) and Support Vector Machines (SVMs). Experimental results demonstrate the effectiveness of the proposed approach, achieving high accuracy in fault classification and early detection. This research highlights the potential of AIdriven acoustic analysis as a non-invasive, cost-effective solution for real-time fault detection in quadcopter propellers, enhancing maintenance efficiency and reducing downtime.

The rapid proliferation of fake news, particularly in low-resource languages like Afan Oromo, poses significant challenges to information integrity and societal trust. This study evaluates the performance of various machine learning models for detecting fake news in Afan Oromo, a language with limited digital resources and linguistic tools. Several supervised learning algorithms, including Support Vector Machines (SVM), Random Forest, Logistic Regression, and deep learning-based approaches such as Long Short-Term Memory (LSTM) networks, are trained and tested on a curated dataset of Afan Oromo news articles. The dataset is preprocessed using tokenization, stemming, and feature extraction techniques tailored to the linguistic characteristics of Afan Oromo. Model performance is assessed using metrics such as accuracy, precision, recall, and F1-score. Results indicate that deep learning models, particularly LSTM, outperform traditional machine learning algorithms in capturing the contextual nuances of Afan Oromo text. However, the study also highlights the challenges posed by the language's morphological complexity and the scarcity of annotated data. This research contributes to the growing body of work on fake news detection in low-resource languages and provides insights into the development of robust models for Afan Oromo and similar languages.

The integration of sound-based artificial intelligence (AI) diagnostics into quadrocopter systems offers a transformative approach for propeller performance optimization. This study explores the use of acoustic data analysis, powered by machine learning algorithms, to detect and diagnose inefficiencies in quadrocopter propellers. By capturing and analyzing sound signatures during flight, the proposed AI-driven system identifies anomalies such as imbalances, wear, and aerodynamic inefficiencies in real-time. The methodology leverages convolutional neural networks (CNNs) and spectral analysis to classify and predict performance degradation, enabling proactive maintenance and optimization. Experimental results demonstrate the system's ability to enhance propeller efficiency, reduce energy consumption, and extend operational lifespan. This research highlights the potential of sound-based AI diagnostics as a non-invasive, cost-effective solution for improving quadrocopter performance and reliability in diverse applications, including aerial surveillance, delivery, and environmental monitoring.

Purpose: Steroid-induced diabetes mellitus (SIDM) is a well-known side effect of corticosteroid agents. While diabetes induced by systemic corticosteroids has been extensively studied for over 60 years, evidence regarding the association between topical corticosteroids and type 2 diabetes remains limited. This study aimed to investigate the relationship between topical corticosteroid use and the risk of type 2 diabetes. Methods: A nationwide, population-based cohort study was conducted using data from the Korea National Health Insurance Service National Sample Cohort Database (NHIS-NSC2). A total of 186,057 adults with new-onset type 2 diabetes were identified as case subjects. Among these, 171,113 adults who had been treated with topical corticosteroids and 14,944 adults with no history of topical corticosteroid use were included and followed prospectively for approximately 9 years. The primary outcome was the incidence of type 2 diabetes following topical corticosteroid use, with additional analyses by frequency and potency of corticosteroid use. Kaplan-Meier curves and Cox proportional hazard models were used to estimate hazard ratios (HR) for type 2 diabetes. Results: The use of topical corticosteroids was significantly associated with an increased risk of type 2 diabetes (adjusted hazard ratio [aHR] 1.13, 95% CI 1.06–1.20). The high-frequency corticosteroid group demonstrated the highest risk (aHR 1.27, 95% CI 1.09–1.48), while the moderate-potency corticosteroid group also showed a significant increase in risk (aHR 1.15, 95% CI 1.04–1.27). Notably, the diabetes risk in the moderate-potency group was comparable to that observed in the systemic corticosteroid group (aHR 1.20, 95% CI 1.14–1.26). Conclusion: In this study, the use of topical corticosteroids was significantly associated with the incidence of type 2 diabetes. These findings highlight the importance of considering diabetes risk factors when prescribing topical corticosteroids.

Porcine epidemic diarrhea virus (PEDV) is a swine enteropathogenic coronavirus causing severe diarrhea and high mortality in neonatal piglets. However, there is little information on monoclonal antibodies (mAbs) against PEDV derived from single B cells of pigs. In this study, we aimed to develop mAbs using antigen-specific single B cell from peripheral blood mononuclear cells (PBMCs) of pigs via fluorescence-activated cell sorting (FACS). Subsequently, the variable region genes of pig-derived mAbs were amplified and cloned into the plasmid pcDNA3.4 bearing the constant region gene of porcine-derived antibody. Pig-derived mAbs were expressed by transfecting the resultant antibody plasmids into HEK293F cells and then were validated. The results showed 21 mAbs were expressed and purified, 20 mAbs of which showed specific binding to PEDV, 19 of which recognized the N protein, and none of which reacted with the S1D protein. 7 out of the 21 mAbs reacted with PEDV HN2021 by Indirect immunofluorescence assay, and western blotting results showed 3 out of the 19 N protein-specific mAbs identified the linear epitopes of the N protein, while the remaining antibodies may recognize its conformational epitopes. This study help develop the diagnostic reagents and antiviral drugs for PEDV.

This paper presents an optimization framework for minimizing both operational and emission-related costs in power grids through demand response (DR) actions, such as load shifting. The proposed approach accounts for the network and location-specific aspects of load, addressing the inherent complexity of bilinear terms in the optimization problem. By applying McCormick relaxation to approximate bilinear interactions between demand shifts and emission factors, we transform the problem into a solvable convex form. A two-step approach is further employed to refine emission values, ensuring accurate emission calculations and optimal demand response strategies. This work contributes to the integration of DR in decarbonized grids, providing a framework for reducing the environmental impact of electricity consumption while maintaining grid stability.

This short communication develops a variational formulation for modeling a protium hydrogen molecular ionization obtained through a high temperature scalar field and an appropriate electric one action. The results are based on standard tools of calculus of variations and optimization theory. The context here addressed is essentially an Euler-Bernoullian one.

A document by Arman Alahyari. Click on the document to view its contents.

”Are infinite computational resources and limitless energy sufficient to achieve lasting peace?” This paper explores this profound question through the lens of quantum-harmonic optimization, presenting the Quantum Harmonic Synchronization (QHS) framework that achieves 99.8% objective alignment while reducing energy waste by 62%. By analyzing 109-dimensional peace-optimization problems, we demonstrate how synchronized quantum gradients and ethical AI architectures can transform resource abundance into sustainable outcomes.

Levinthal’s Paradox poses a fundamental challenge in protein folding: how do proteins like cytochrome c (104 residues) achieve their native states in milliseconds despite 3^104 ≈ 10^49 possible conformations? The Free-Energy Principle (FEP) models adaptation via energy minimization but lacks the kinetic efficiency to explain rapid folding. We introduce the Chaos-Based Useful Information Principle (UIP), an extension of FEP, integrating chaotic dynamics to maximize goal-directed information (U = I(s,r) + C(s,r) - D_KL) and accelerate uncertainty reduction. UIP leverages chaos (lambda^+ ≈ 0.09) to collapse conformational spaces, validated by simulating cytochrome c folding, reaching the native state (z = 4.0) in 15 ms, consistent with experimental data (10–20 ms; Sosnick et al., 1994). The Rössler-based model aligns with real kinetics, with chaotic dihedral dynamics (x-y trajectory) driving efficiency. UIP builds on FEP’s thermodynamic foundation—mapping internal to external states—while overcoming its slow, reactive nature with proactive exploration. This resolves Levinthal’s Paradox, surpassing FEP and other models (e.g., folding funnels, nucleation-condensation). Physiologically, chaos mirrors cellular perturbations, offering insights into folding mechanisms and diseases (e.g., amyloidosis). UIP’s FEP roots ensure robustness, while its innovation enhances kinetic prediction, suggesting experimental validation via single-molecule techniques. This framework bridges structural biology and dynamic physiology, with implications for protein design and therapeutic strategies.

The insect olfactory system employs a diverse array of olfactory-related proteins to facilitate odor detection and signal transduction. Their expression and regulation are crucial for mediating communication between themselves and environments. Lytta sifanica (Coleoptera: Meloidae) is one of the most important economic pests and is famous for producing a toxic substance cantharidin. However, the genes underlying olfactory sensation are lacking in this Blister Beetle. In this study, the transcriptomes of adult L. sifanica antennae were sequenced and analyzed. A total of 70 chemosensory genes, including 17 odorant binding proteins (OBPs), 5 chemosensory proteins (CSPs), 13 gustatory receptors (GRs), 17 odorant receptors (ORs), 13 ionotropic receptors (IRs) and 5 sensory neuron membrane proteins (SNMPs) were identified based on sequence homology analysis and phylogenetic reconstruction. The expression patterns of all candidate genes in the antennae of adults, head, mouthparts, pronotum, foreleg tarsus, abdomen skin, wings were confirmed by RT-PCR. The analyses demonstrated that all protein families related to olfaction are widely expressed in examined tissues. However, the expression patterns varied for different families. Eleven members are predominantly expressed in the antennae (Lsif_OBP1/2/19d, Lsif_OR2/20/49b, Lsif_IR2a, Lsif_GR7/127, Lsif_CSP1, Lsif_SNMP2/4). Twenty-two members were exclusively expressed in the mouthparts (Lsif_OBP70/56d2, Lsif_GR12a/21/28a/68a), foreleg tarsus (Lsif_OR6/67c, Lsif_GR12, and Lsif_OBP2) and were abundant in the non-olfactory tissues head (Lsif_OBP99a, Lsif_OR9a/49b, Lsif_IR25a/56e, and Lsif_CSP6), pronotum (Lsif_OBP5/C20, Lsif_Orco3, Lsif_OR13, and Lsif_IR7), abdomen skin (Lsif_SNMP5), suggesting their various functions in the olfactory system of L. sifanica. This research offers an extensive resource for investigating the chemoreception mechanism in beetle L. sifanica.

Effective conservation requires effective biodiversity monitoring. The pace of global biodiversity change far outstrips the ability of manual fieldwork to monitor it. Therefore, technological solutions, like camera traps, have emerged as a crucial way to meet biodiversity monitoring needs. Camera traps produce vast amounts of data and so AI is increasingly used to label images with species identities. However, AI struggles to identify species from new locations that are not part of the training data (‘generalisation’). Resolving this is crucial for the promise of automated biodiversity monitoring to be realised. Here we use ‘copy-paste’ augmentation to help resolve the generalisation challenge. Copy-paste augmentation refers to isolating animal ‘segments’ from existing images and pasting the segments onto novel backgrounds, to create new, synthetic images that are then used as part of the training data. Theoretically, this could make a model agnostic to backgrounds and therefore more able to generalise to unseen locations. While generation of synthetic images is commonly used as an augmentation method in other fields, such as medicine, it has not been used before in biodiversity science. We found that copy-paste augmentation improved the ability of AI to identify species in new, unseen locations by $8\pm2\%$. There was species-level variation in improvement, but the vast majority of species benefited from the approach. We found mixed results when using copy-paste augmentation on models trained with very small numbers of images (1-8 per species). Copy-paste augmentation improves the ability of AI models to generalise to new, unseen locations. Our method also shows promise for resolving the challenge of long-tailed camera trap data.

Reproductive barriers limit gene flow and drive population divergence. For internal fertilizers, sperm morphology plays an important role in reproductive barriers, as successful fertilization depends on how well sperm perform in the female environment. Specifically, sperm length must be adapted to fit the female reproductive tract and storage organs. In species where sperm competition occurs, i.e. where multiple males compete to fertilize a female’s eggs, selection pressure can favor an optimal sperm length, reducing variation over time. Additionally, variation in sperm length may result in different optima in locally adapted populations or ecotypes that may facilitate further divergence. We investigated sperm length in species and ecotypes of Littorina, a genus of promiscuous marine snails with internal fertilization. Various Littorina species mate inter-specifically, where sperm length differences might prevent hybridization across species borders. Additionally, several Littorina species have ecotypes adapted to different shoreline environments where reproductive barriers like sperm length divergence might play a role in reinforcing these barriers. Due to their promiscuity, sperm competition probably plays a role in Littorina, and sperm length variation can give insights in sperm competition intensities. Littorina snails have two types of sperm, eusperm which is the fertilizing sperm, and parasperm. This study examined eusperm length in four species (L. fabalis, L. littorea, L. obtusata, L. saxatilis), and two different ecotypes in both L. fabalis and L. saxatilis. The ecotypes of both L. fabalis and L. saxatilis differed in eusperm lengths, suggesting that this trait may be involved in prezygotic reproductive barriers between ecotypes of these species. Among-species differences in eusperm length variation were observed and may be a result of different sperm competition intensities.

The study investigated the spatial distribution of Straw-Colored Fruit Bat (Eidolon helvum) roosts within the Obafemi Awolowo University (OAU) campus in Nigeria. Using field surveys and GIS analysis, the researchers mapped bat roost distribution, identified hotspots, and explored tree species preferences. Key findings include pronounced roosting clusters, particularly in the Faculty of Science and Faculty of Administration areas, with Celtis zenkeri identified as the most preferred roosting tree species. The study emphasizes the importance of specific tree attributes, such as height and canopy area, in roost site selection within the urban campus environment. The research contributes to a better understanding of bat roosting ecology and provides insights to guide conservation strategies, recommending that urban planners prioritize the preservation of preferred tree species like Celtis zenkeri.

Cytochrome P450 (CYP)3A is involved in the metabolism of more than 50% of prescribed drugs, and reports indicate that 70% of CYP3A inducers are also CYP3A inhibitors, thereby complicating the interpretation of induction based DDI potential and reducing the confidence in clinical outcome extrapolation for co-regulated enzymes such as CYP2C8, 2C9 and 2C19. While the preclinical assessment of CYP3A4 induction has been conducted extensively, the evaluation of CYP2C induction is hindered by low dynamic response in standard monoculture hepatocyte models, which may prompt additional clinical DDI investigations. This study was aimed to characterize the induction potential of CYP2C enzymes in all human hepatocyte triculture model (HTC), a two-dimensional hepatic system with human primary hepatocytes, stromal, and endothelial cells. The in vitro induction potential of known inducers of CYP2C8, CYP2C9, CYP2C19 and CYP3A was assessed using the HTC model. In addition, hepatocytes from a single donor were plated in the sandwich culture (SC) model for direct comparison of induction endpoints and the induction parameters were used for simulating clinical PK implications. RNA-seq results showed distinct basal expression differences in CYPs, transporters and transcription factors between both models, potentially suggesting better recapitulation of native liver hepatocytes in the HTC model. Compared to SC, the HTC model showed robust induction of CYP2C. By incorporating the induction parameters obtained from the HTC model into PBPK models, an excellent correlation was obtained relative to clinical outcomes for CYP2C8, CYP2C9, CYP2C19 and CYP3A4. Overall, this study provides a potential approach to quantitatively assess CYP2C-induction risk preclinically.

Background: Antimicrobial resistance (AMR) has risen as a global health problem. In low-income countries, AMR individual testing is costly and often not done. Studies elsewhere, have shown that clinical wastewater is a single sample that represent a pool of so many individuals and comprise a cost effective screening endeavor that can improve empirical prescription (which is otherwise associated with exacerbation of AMR). This cross sectional study aimed at isolating and profiling AMR bacteria from hospital wastewater of Soroti regional referral hospital (SRRH). Methodology: In this cross section study, we prospectively collected samples of hospital wastewater over a period of one month from different hospital manholes. We cultured samples on different selective media and pure plated representative colonies depending on the colony characteristics. We then tested the isolates for antibiotic resistance using the Kirby-Bour disk diffusion method. We interpreted results according to CLSI 2020 using the diameters of zones of inhibition. We defined multidrug resistance as non-susceptibility to at least one agent in three or more antimicrobial categories. Results: Taken together, we isolated 35 bacteria from 22 samples collected from different manholes of the SRRH. Klebsiella spp were the majority (15/35; 43%). This was followed by Bacillus cereus (9/35: 26%), Escherichia coli (7/35; 20%), unidentified coliform 3 (3/35; 8%) and Pseudomonas (1/35; 3%). All isolates of Klebsiella spp were resistant to Ampicillin and Piperacillin/Tazobactum. Resistance to Cefotaxime and Ciprofloxacin was (12/15; 80%). However, Klebsiella was most susceptible to Gentamycin and Azithromycin (13/15; 86.7%). Susceptibility to Meropenem and Amikacin was (12/15; 80%). Other drugs that we tested for resistance against Klebsiella included Norfloxacin, Doxycycline, and Tetracycline (11/15; 26.7%), (9/15; 60%), (6/15; 42.9%) respectively. All isolates of E. coli were resistant to Cefotaxime, Ampicillin, Cefuroxime and Tetracycline. The isolates were most susceptible to Meropenem and Amikacin (1/7; 14.3%) and (2/7; 28.6%), respectively. They were highly resistant to Nalidixic Acid, Ciprofloxacin, Piperacillin/ Tazobactum and Levofloxacin (6/7; 85.7% for each drug). Other drugs that we tested for resistance included; Azithromycin, Norfloxacin and Doxycycline with resistance of (5/7; 71.4%) and Gentamycin with (4/7; 57.1%).

A document by Peter Mounas. Click on the document to view its contents.

Forest conversion into agriculture is a major driver of biodiversity loss in the tropics. In West Africa, Guinea-Bissau lost two thirds of its dense forests over the last two decades due to the unprecedent expansion of cashew orchards. However, the effects of this land-use change on biodiversity remain poorly understood. To address this gap, we examined small mammal species richness, abundance, and composition across 24 sites evenly distributed between forests and cashew orchards in Cantanhez National Park, Guinea-Bissau. Small mammals were live-trapped, molecularly identified, and their diversity related to local (i.e., floor and understorey layers obstruction, canopy openness, palm and liana density, maximum tree height, and tree density and richness) and landscape-scale variables (forest and cashew cover, and edge density). Based on 5,760 trap-nights, we recorded 105 individuals from five rodent and two shrew species. All species were recorded within cashew orchards, while only three were found in forests. Praomys rostratus, the species with the highest number of individuals (n = 72), was found almost exclusively in forests. Species richness increased with understorey obstruction, species abundance decreased with cashew cover, and species composition varied with forest cover, tree height, and canopy openness. Overall, our findings reveal a critical link between both local and landscape-scale variables and small mammal assemblages. Cashew expansion is driving the decline of the forest-dependent P. rostratus, while benefiting generalist and open-area species. We recommend regulating the expansion of cashew orchards to guarantee the persistence of forest-dependent biodiversity in the African tropics.

This study was conducted at Ferlo (Sénégal) and Mankarga (Burkina Faso) natural rangelands, to investigate the woody plant diversity, vegetation structure, and regeneration status of the rangelands. A systematic sampling design was employed to collect vegetation data. 64 quadrats of varying size were laid in each of the rangeland. The inventory unit was 1000 m² (50 m x 20 m) in savannas and 500 m² (50 m x 10 m) in gallery forests. Data collected from each quadrat included the Diameter at breast height (DBH), the height, the regeneration, and woody species density. The DBH was estimated for woody plant with a DBH greater than 5 centimetre. Shannon-Wiener diversity, equitability indices, and the structural analysis was carried out based on frequency, density, DBH, height and basal area. The Importance Value Index (IVI) was also computed. The regeneration status of the studied rangelands was evaluated based on the number of seedlings. A total of 89 fodder tree species belonging to 64 genera and 25 families from Ferlo and 80 species belonging to 57 genera and 23 families from Mankarga was identified. Fabaceae and Combretacea was the dominant family in the two rangelands. Analysis of selected woody species showed diverse population structures. The findings of this study revealed that small trees dominated the rangeland suggesting its status under a secondary stage of development. Some woody species urgently need conservation. Therefore, local and regional stakeholders must collaborate to develop and implement sustainable rangeland management and conservation strategies Keywords: Woody species, Conservation; Ferlo rangeland; Mankarga rangeland; Senegal; Burkina Faso