Browse Articles

Predicting COVID-19 severity in pediatric patients using machine learning: a comparative analysis of algorithms and ensemble methods

Abstract COVID-19 has posed a significant global health challenge, affecting individuals across all age groups. While extensive research has focused on adults, pediatric patients exhibit distinct clinical characteristics that necessitate specialized predictive models for disease severity. Machine learning offers a powerful approach to analyzing complex datasets and predicting outcomes, yet its application in pediatric COVID-19 remains limited. This study evaluates the performance of machine learning algorithms in predicting disease severity among pediatrics. A retrospective analysis was conducted on a dataset of 588 pediatric with confirmed COVID-19, incorporating demographic, clinical, and laboratory variables. Various machine learning models were trained and assessed, with a SuperLearner ensemble model implemented to enhance predictive accuracy. Among the models, Random Forest exhibited the highest performance, achieving an accuracy of 90.1%, sensitivity of 90.2%, and specificity of 90.1%. The SuperLearner ensemble further improved predictive performance, demonstrating the lowest mean risk estimate. Key predictors, including oxygen saturation, respiratory parameters, and specific laboratory markers, played a crucial role in distinguishing severe from non-severe cases. These findings emphasize the potential of machine learning, particularly ensemble methods, in improving risk stratification for pediatric COVID-19. Integrating these predictive models into clinical practice could support early identification of high-risk patients and optimize clinical decision-making.

Isoform analysis of heterozygous putative splicing variants at the allele level using nanopore long-read sequencing

Abstract One of the challenges in clinical genetics for rare diseases and personalized medicine is evaluating isoform alterations arising from heterozygous putative splicing variants at the allele level. Our aim was to analyze these variants by dividing cDNA or direct RNA nanopore long reads into two alleles, referencing whole-genome sequencing data containing allele-informative single nucleotide variants and then comparing the allele-separated reads using Full-Length Alternative Isoform analysis of RNA (FLAIR), a previously published bioinformatics tool for isoform analysis. In this study, we developed an allele-separative bioinformatics pipeline and described its performance. We applied our pipeline to previously published nanopore direct RNA sequencing data, as well as 5’ cap-trapping full-length cDNA nanopore sequencing (CTR-seq) data from blood samples of three individuals. We successfully identified heterozygous splicing variants associated with significant isoform differences between alleles. Furthermore, we uncovered the effects of a novel pathogenic splicing variant in PYGM on isoforms in a compound-heterozygous case of McArdle disease using nanopore cDNA amplicon and targeted genomic sequencing. This study demonstrates the utility of nanopore long-read sequencing for isoform analysis at the allele level, providing a valuable approach to evaluating the direct consequences of heterozygous splicing variants in individuals.

Temperature dependent microstructural defects and surface charge effects on antioxidant activity of green synthesized nanoceria

CB2 and TRPV1 receptors in inflammatory state of macrophages from sickle cell anemia pediatric/young adults

Abstract Sickle Cell Disease (SCD) is a monogenic disorder characterized by the production of abnormal hemoglobin. Polymerization of HbS causes sickling of red blood cells (RBCs) evidenced by acute adverse events and persistent inflammatory state, vasculopathy and organ damage. Sickled RBCs cause an anemic condition and vaso-occlusive crisis which trigger leukocytes, endothelial cells, and platelets. Due to these events, SCD patients unveiled an elevated level of pro-inflammatory cytokines, which contribute to the ongoing inflammatory state, oxidative stress, and other severe complications. SCD patients also experience neuropathic, inflammatory, and nociceptive pain. The discovery of novel therapeutic approaches and targets to counteract and manage inflammation in SCD are needed. Our study aimed to better understand the role of macrophages in SCD inflammation by first investigating their phenotype and then studying the iron metabolism involvement in the inflammatory processes. Therefore, given the importance to find novel therapeutic approach to contain and manage inflammation in these patients, and considering the role of CB2 and TRPV1 in this process, we decided to investigate the expression of these receptors and the effects of their stimulation on inflammatory state in SCD macrophages.

Carnivorous plants can decompose the polyesters poly(ethylene terephthalate) and poly(butylene adipate terephthalate)

Developing real-time IoT-based public safety alert and emergency response systems

Land use classification using multi-year Sentinel-2 images with deep learning ensemble network

Abstract Accurate land use classification is essential for urban planning, environmental monitoring, and agricultural management. Sentinel-2 satellite imagery provides rich spatial and spectral information suitable for this purpose. This study proposes a deep learning ensemble network named IRUNet, which integrates InceptionResNetV2 with a UNet framework for multi-year Sentinel-2 imagery classification over the Katpadi region (2017–2024). Unlike prior works, IRUNet utilizes multi-scale feature fusion and incorporates Test-Time Augmentation (TTA) to enhance prediction robustness. While the data spans multiple years, each year is treated as an independent input without modeling temporal sequences. The proposed method demonstrates superior performance over UNet, ResUNet, and Attention-UNet models, achieving an accuracy of 98.21% and Dice similarity coefficient (DSC) of 88.96%. Additional metrics including precision (94.71%), recall (89.19%), F1-score, and Kappa coefficient have been reported. This research contributes a high-performance, generalizable framework for multi-year land use classification.

Cyanobacterial bloom causes expansion of isotopic niche areas and overlap in crustacean zooplankton

Abstract We aimed to study how cyanobacterial blooms affect the use of the basal resources by three groups of crustacean zooplankton (calanoid and cyclopoid copepods, Daphnia spp.). We used measurements of naturally occurring stable isotopes of carbon (δ13C) and nitrogen (δ15N) to quantify the areas of isotopic niches (sample size-corrected standard ellipse areas; SEAc) of planktonic crustaceans during the pre-bloom and cyanobacterial bloom phases. In the pre-bloom phase, SEAcs accounted for 15.0‰2 in calanoid copepods, 21.2‰2 in cyclopoid copepods and 14.4‰2 in Daphnia spp. During the cyanobacterial bloom phase, the SEAcs of studied animals increased to 37.8, 27.0 and 43.6‰2 respectively. In addition, the overlap among the niches of the crustacean groups increased during the bloom phase compared to the pre-bloom phase. The results suggest that, despite reduced diversity of basal resources during the cyanobacterial bloom, crustaceans exhibited dietary adaptability. This involved a shift toward alternative food sources.

Inverse unit compound Rayleigh distribution: statistical properties with applications in different fields

Investigation into the supply-demand relationship of carbon sequestration in the yellow river basin using the optimal parameter geographical detector model

Genetic evidence reveals phosphatidylcholine as a mediator in the causal relationship between omega-3 and multiple myeloma risk

Comprehensive characterization and extraction implications of ion adsorption rare earth deposit from a South American source

Effect of rGO synthesized from different precursors on the enhancement in mechanical properties of GFRPs

Cancer stem cells and Lon-noncRNA promotes invasion, metastasis and tumor growth in breast cancer through regulation of signaling pathway

Abstract Breast cancer (BC), the most common malignant tumor in women, continues to be a leading cause of cancer-related deaths globally. A major challenge in managing BC, especially in metastatic cases, is the lack of reliable early diagnostic biomarkers. Metastatic breast cancer stem cells (MBCSCs) play a critical role in tumor progression, resistance to therapy, and disease recurrence. This study aimed to explore the molecular pathways connecting the long non-coding RNAs (lncRNAs) HOTAIR, UCA1, and MALAT1 with breast cancer stem cell-related genes FOXC2, SNAIL, and ZEB, focusing on their involvement in transcriptional regulation, proliferation, and survival. Peripheral blood samples and plasma were collected from 30 women diagnosed with metastatic breast cancer (MBC, stage IV) and 30 healthy controls. Gene expression levels were measured using quantitative real-time PCR (qRT-PCR). Our findings revealed a significant upregulation of SNAIL and FOXC2 in MBC patients compared to healthy controls (p < 0.001). The median expression levels of SNAIL (16.4) and FOXC2 (19.5) were substantially higher in the metastatic group than in healthy individuals (SNAIL: 6.42, FOXC2: 7.23). Conversely, the expression levels of HOTAIR, UCA1, MALAT1, and ZEB did not show statistically significant differences between the two groups (p > 0.05). Correlation analysis indicated a strong positive association between FOXC2 and SNAIL expression (r = 0.41), suggesting a potential shared functional role in disease progression. These results suggest that SNAIL and FOXC2 could serve as potential prognostic biomarkers in MBCSCs, whereas HOTAIR, UCA1, MALAT1, and ZEB may not independently predict metastasis or survival outcomes. Further research is necessary to explore the therapeutic implications of these genes in metastatic breast cancer.

Early changes in corticospinal excitability for subliminally presented fearful body postures

Abstract Fearful body expressions convey critical information that is rapidly and preferentially processed, facilitating swift motor reactions to potential dangers. Consistent evidence has shown that even the subliminal presentation of fear-related expressions can impact visual processing and autonomic responses, increasing sensory vigilance for monitoring potential threats. However, it remains unclear whether the presentation of non-visible emotional bodies modulates corticospinal excitability (CSE) in the observer. To investigate this, we asked 22 healthy participants to perform a sex discrimination task involving neutral target body postures, preceded by the brief subliminal presentation of fearful, happy, or neutral body postures. CSE was tested using Transcranial Magnetic Stimulation (TMS) at early time points (70, 90, and 110 ms) after target stimulus onset. Results showed a significant CSE reduction in the dominant hemisphere for subliminal fearful primes compared to happy and neutral primes. This CSE suppression was independent of the time of stimulation, participants’ subjective or objective awareness, metacognitive sensitivity, or personality traits. Our findings highlight an early automatic activation of the motor system in response to subliminal fearful stimuli, supporting the view that fearful expressions, even when not consciously perceived, activate basic survival mechanisms for monitoring and preparing fast motor responses to potential threats.

Propensity score matching analysis of the effect of four or more antenatal care visits on basic childhood immunization in Ethiopia

Association between social determinants of health and systemic lupus erythematosus: a nationally representative analysis of 2017–2021 data

Abstract Limited US recent data is available on the prevalence of systemic lupus erythematosus (SLE) by patient’s social determinants of health (SDOH). Careful assessment of individual SDOH that affects SLE is crucial, as such evidence could help improve care and hence reduce health disparities for patients with SLE, especially for those who are most vulnerable and at the highest risk of poor outcomes. We estimated the prevalence of systemic lupus erythematosus (SLE) overall and by patient’s social determinants of health (SDOH), and also explored the associations between SDOH and SLE. We conducted a population-based cross-sectional study using Medical Expenditure Panel Survey 2017–2021 data. Patients with SLE were those with both SLE diagnosis and either had SLE-related medication use and/or visited a rheumatologist in the survey year. SDOH domains included economic stability, education, healthcare access and quality, social and community context, and neighborhood and built environment. Average annual prevalence of SLE by SDOH was determined. Separate logistic regressions were used to examine the association between each SDOH and SLE, controlling for confounders. Average annual SLE prevalence was 199 per 100,000 US adults (95% confidence intervals:170–224). In the economic stability domain, those with low family income showed higher odds of SLE than those with high family income (adjusted odds ratio (AOR) = 2.779, p < 0.05). Within the social and community context and neighborhood and built environment domains, non-Hispanic Black patients (AOR = 2.429, p < 0.05) and patients with any psychological distress (AOR = 2.127, p < 0.05) had higher odds of SLE, than their respective counterparts, respectively. Within the healthcare access and quality domain, those with Medicaid insurance had higher odds of SLE (AOR = 2.540, p < 0.05) than those with private insurance. Also, patients in the highest SDOH burden quartile had higher odds of SLE (AOR = 2.039, p < 0.05) than those in the lowest SDOH burden quartile. We identified several subgroups of patients especially those with higher social disadvantage and a higher SDOH burden. The prevalence of SLE increased with a higher SDOH burden.

Complex linear Diophantine fuzzy Dombi prioritized operators-based MULTIMOORA approach with applications to sustainable energy planning

Mathematical modeling of tumor-immune dynamics: stability, control, and synchronization via fractional calculus and numerical optimization

Obesity in chronic spinal cord injury is associated with poorer body composition and increased risk of cardiometabolic disease

Abstract With increased longevity after spinal cord injury (SCI), cardiovascular disease has emerged as a major cause of morbidity and mortality. We evaluate the association of body composition and injury level with cardiometabolic disease (CMD) risk factors. Sixty-two individuals (69% male, 31% female) with chronic SCI (mean duration 7.4, SD ± 5.8 years) were recruited. Mean age of participants was 34.4 ± 12.1 years with BMI of 23.8 ± 5, while 64% had BMI < 25 and 11% >30. Total and percent truncal fat correlated positively (p < 0.05) with serum triglycerides, non-high-density lipid cholesterol, c-reactive protein (CRP), oral glucose tolerance test (OGTT), and measures of insulin resistance. Those with obesity in SCI (defined as BMI ≥ 22) had increased total and trunk mass and fat percentage, unfavorable lipid profiles and evidence of insulin insensitivity. Total fat was associated with CMD risk factors, including insulin resistance (OGTT 60 min r = 0.47, p < 0.05; homeostasis model assessment [HOMA] r = 0.62, p < 0.05), serum triglycerides (r = 0.31, p < 0.05), and inflammation (CRP r = 0.43, p < 0.05). Obesity in SCI related to higher CMD risk, while time since injury and injury level (paraplegia versus tetraplegia) did not. Future studies may evaluate roles of nutrition, exercise, sleep-promotion, and pharmaceuticals to lower neurogenic obesity and chronic CMD risk.