Background: Alström syndrome (ALMS) and Bardet-Biedl syndrome (BBS) are rare ciliopathies characterized by multisystem involvement, including obesity, insulin resistance, and type 2 diabetes. Systemic metabolic dysfunction may influence the oral microbiome; however, integrative analyses that combine microbial and metabolic profiles in these disorders remain limited. Methods: Saliva and gingival crevicular fluid (GCF) samples were collected from genetically confirmed ALMS and BBS patients, as well as from obesity and healthy control groups. Microbial communities were profiled using V3-V4 16S rRNA gene amplicon sequencing, and untargeted metabolomic profiling was performed by gas chromatography-mass spectrometry. Microbiome-metabolome associations were evaluated using Spearman's rank correlation analysis, followed by multi-omics integration using Multiple Co-Inertia Analysis (MCIA) and the supervised Data Integration Analysis for Biomarker discovery using Latent cOmponents (DIABLO) framework (mixOmics). Results: Integrated analysis identified distinct microbiome-metabolome association patterns in ALMS and BBS. Compared with controls, the ALMS+BBS group showed enrichment of Prevotella, Enterococcus, and Eikenella, alongside reduced Lactobacillus abundance. Metabolomic profiling revealed alterations in amino acid, fatty acid, and carbohydrate metabolism. GCF exhibited structured associations between metabolites and Firmicutes, Proteobacteria, and Actinobacteriota, whereas saliva showed broader interaction networks. These associations were absent or markedly weaker in obesity and healthy controls. MCIA demonstrated coordinated variation across the oral microbiome, salivary metabolome, and GCF metabolome, while DIABLO identified a shared multi-omics signature. Conclusions: Coordinated shifts in amino acid, lipid, and central carbon metabolism can be linked to oral microbial reorganization in ALMS and BBS. Integrative multi-omics analyses identified coordinated microbiome-metabolome signatures across the oral microbiome, saliva, and GCF. These findings warrant validation in larger longitudinal and functional studies.
“Mirror life”, self-replicating organisms composed of nonnatural-chirality biomacromolecules, presents a future threat with potentially global consequences. Consequently, there is strong agreement among experts that it should not be created. However, there is some disagreement over how effective existing medical countermeasures might prove against mirror bacteria, in the event that they were created. Here, we leverage computational chemistry methods including docking and molecular dynamics to determine the likely binding efficacy of existing antibiotics against natural and mirror bacterial protein targets. We find that most existing antibiotics fail to bind to mirror bacterial protein targets, unlike their natural-chirality targets. This suggests altered binding of current medical countermeasures, which may impact antimicrobial activity against mirror bacteria if the latter were created.
Introduction: Electrical impedance tomography (EIT) is a noninvasive, radiation-free imaging modality that provides real-time information on regional lung ventilation from wearable sensors. Chronic obstructive pulmonary disease (COPD) is a highly prevalent respiratory disease causing persistent and progressive airway obstruction. The diagnosis is based primarily on spirometry, in which a Tiffeneau index (forced expiratory volume in 1 s/forced vital capacity ratio) below 0.7 is a key feature of the definitive diagnosis. Despite many advances in medicine, there is a lack of widely available methods for estimating airway obstruction using noninvasive bedside measurements, which could facilitate assessment in patients who are unable to perform standard spirometry, e.g., patients after laryngectomy and those with chronic tracheostomies. Moreover, the prevalence of COPD in this group of patients may be substantial. Methods: The study examined the relationship between EIT-derived signal features and the Tiffeneau index through a comprehensive statistical and machine learning analysis of patient data collected using the Dräger EIT system. Data from 15 adult patients who successfully completed conventional spirometry and EIT measurements were analyzed. Patients conducted the examination a couple of times; consequently, 30 measurements were collected. A total of 755 time-series features, complemented by physiological measurements, were extracted, analyzed, and evaluated using correlation metrics, P-value testing, categorical associations, and Shapley additive explanations explainability. Results: Frequency-domain features (fast Fourier transform angle coefficients), entropy measures, and autocorrelation-based descriptors have the strongest associations with the Tiffeneau index. Feature-selected machine learning models demonstrated that EIT-derived time-frequency features, combined with anthropometric variables (weight and body mass index), could approximate Tiffeneau index values in this small cohort (best model: R 2 = 0.72 on the testing set). Conclusion: These findings support the exploratory feasibility of EIT-based, noninvasive approaches for pulmonary function estimation in a general clinical cohort and motivate future validation studies in patients for whom spirometry cannot be performed.
Foundation vision encoders are rapidly emerging as the standard for retinal artificial intelligence. Yet, ophthalmology still lacks a comprehensive benchmark, leaving model selection for basic science and clinical translation as guesswork. Here, we present a large-scale comparison of 34 pretrained encoders on 39 classification tasks covering color fundus photography, optical coherence tomography, scanning laser ophthalmoscopy, and ultrawidefield imaging. Using a unified pipeline, we compare frozen-feature evaluation, linear probing, and end-to-end fine-tuning to determine which models translate into strong downstream performance. We show that ophthalmic transfer is highly task dependent: no single encoder dominates, and model rankings vary across datasets. Contrary to common expectations, retina-specific pretraining does not confer an advantage. Instead, several natural-image and cross-domain medical encoders match or surpass ophthalmology-specialized models, with the histopathology-pretrained Virchow achieving the strongest overall performance. In addition, pathology-pretrained encoders consistently place near the top, revealing the value of cross-domain pretraining for ophthalmic applications. We further show that inexpensive proxy evaluations are unreliable substitutes for full fine-tuning. Across fairness analyses, all encoders exhibit similar age- and sex-associated performance gaps, and larger models appear more sensitive to suboptimal learning rates, whereas smaller encoders are robust. Together, these findings provide an objective reference for encoder selection in ophthalmology and show that reliable retinal artificial intelligence depends not only on model scale or domain-specific pretraining but also on careful, protocol-aware evaluation. By releasing our code, splits, and benchmarking pipeline, we aim to establish a transparent foundation for future ophthalmic foundation-model research.
Many problems in biomedicine can be posed as binary classification. When they are addressed using artificial intelligence methods, though, average performance alone does not show whether a dataset is artificial intelligence ready, whether the endpoint is clinically valid, or whether errors are unevenly distributed across patient subgroups. This article presents the Fairness-Aware Data Orchestration Pipeline (FADOP), a reusable workflow that analyzes biomedical datasets, trains baseline binary classifiers, audits subgroup error patterns, tests mitigation strategies, and generates a documented recommendation. Such a pipeline is intended for systematic evaluation before clinical translation, not as an automatic deployment tool. Two publicly available case studies illustrate its use: the HIV-related ACTG175 dataset was repurposed from a treatment-comparison trial into a 1-year baseline mortality-prediction task, with death by day 365 as the positive class rather than the cid AIDS/failure composite endpoint; then, a stroke-risk dataset was analyzed as direct event prediction. The case studies show how the same workflow can generate cohort, performance, fairness, mitigation, and recommendation evidence across different rare-event clinical datasets.
Cell populations grow, shrink, and reshuffle their composition as individual cells divide, arrest, and die. Because cell division is paced by progression through the cell cycle, cell-cycle control is the fundamental axis linking intracellular regulation to population-level dynamics. Quantitatively capturing how cell-cycle regulation affects population dynamics is thus central to understanding tissue homeostasis, tumor expansion, immune responses, and the performance of cell-based bioprocesses—from tissue engineering and regenerative applications to the manufacturing of vaccines and recombinant therapeutics. In this view, a broad spectrum of mathematical models has been proposed to describe cell population dynamics, ranging from phenomenological growth laws that treat net proliferation as a black box to structured descriptions that resolve single-cell heterogeneity and link population change to cell-cycle progression. In this review, we survey deterministic frameworks for modeling cell-cycle-informed population dynamics, highlighting how modeling choices map onto accessible experimental readouts and biomedical and biotechnological questions. In doing so, we aim to provide a theoretical roadmap for readers new to the field who seek to translate intracellular cell-cycle regulation into empirically grounded population-level models. We organize models along 2 conceptual dimensions: the representation of cell-to-cell heterogeneity through increasing levels of population structure and the level of mechanistic specification of cell division through cell-cycle regulation. In particular, we discuss practical approaches to couple cell-cycle regulation to population-level dynamics—from coarse-grained to explicit multiscale couplings—and how external perturbations enter such couplings. We conclude by outlining open challenges toward cell-cycle-aware population models that match mechanistic resolution with experimental identifiability.
New approaches to drug discovery are urgently needed, especially for diseases with complex molecular mechanisms where current treatments show limited efficacy. We present a data-driven workflow for drug repositioning that integrates high-throughput and virtual screening, molecular docking, and mechanistic pharmacokinetic modeling with biological validation using lung fibrosis as a model disease. High-throughput screening assays and LINCS L1000, SwissTarget, and Unified Knowledge Space (UKS) databases identified ouabain and helenalin as candidates targeting lung fibrosis-related genes. Pharmacophore modeling, docking analysis, and physiologically based pharmacokinetic modeling confirmed drug-like properties comparable to current lung fibrosis treatments. In silico findings were validated using RNA-sequencing data from idiopathic pulmonary fibrosis patients and quantitative polymerase chain reaction data from human alveolar epithelial cells exposed to profibrotic transforming growth factor-β. In vitro, helenalin and ouabain induced distinct gene expression profiles. The anti-inflammatory signature observed for helenalin points to preliminary functional potential warranting further investigation. This workflow can be generalized across diverse therapeutic areas to accelerate cost-effective drug discovery and repositioning.
Type 2 diabetes mellitus (T2DM) and metabolic syndrome (MetS) in children and adolescents are characterized by altered lipid metabolism and gut microbiota. Prior untargeted (nonquantitative) lipidomics of the present T2DM and MetS pediatric cohort revealed alterations in plasma lipids. Fatty acids (FAs) bonded to plasma phospholipids (PLs) and cholesterol esters (CEs) reflect endogenous metabolism and exogenous sources. The present study focused on targeted quantification of FAs esterified to plasma PLs and CEs in children and adolescents with T2DM and MetS and healthy controls ( n = 60, ages 7 to 17). Regression models and Spearman correlations assessed associations of esterified FA with the disease, metabolic risk factors, pro-inflammatory cytokines, and gut microbiota composition. T2DM and MetS groups featured higher concentrations of saturated FAs esterified to PLs (C17:0) and CEs (C10:0, C12:0, and C24:0) than healthy controls. Also, both groups had higher omega-6 FA levels, including dihomo-γ-linolenic acid (C20:3n-6) in both plasma fractions (PLs and CEs), C22:5n-6 in PLs, and C18:2n-6 and C20:4n-6 in CEs. These FAs were inversely correlated with high-density lipoprotein cholesterol and directly correlated with obesity, triglycerides, and insulin resistance. Also, MetS had a high CE-omega-6/omega-3 ratio. Gut microbial taxa associated with T2DM and MetS after Tanner, sex, and body mass index percentile adjustment were Agathobacter , Dorea , Fusicatenibacter , and Gemmiger , with higher abundances than those in healthy controls, and the genus Faecalimonas at lower abundances. This work contributes to the current knowledge of lipid metabolism and the role of gut microbiota in T2DM and MetS in children and adolescents.
Polymerase cycling assembly (PCA) allows gene synthesis from the assembly of overlapping oligonucleotides. For optimal yield, the sets of overlapping fragments should present uniform hybridization temperatures and minimal spurious dimer formation, conditions that become particularly difficult to fulfill for short, GC-biased, or dimer-prone sequences. We present a new algorithm for designing overlapping oligonucleotides for PCA. Unlike heuristic-based methods, the approach maps the design onto a statistical physics model whose low-temperature solution can be obtained exactly, through a transfer-matrix formalism, when spurious heterodimer interactions are negligible; when they are not, it is complemented by a simulated-annealing repair step and, optionally, by codon redesign, generating high-quality, thermodynamically consistent oligonucleotide sets. We evaluated the method on over 10,000 sequences previously deemed unsuitable for synthesis, successfully recovering roughly 80% of them, and more than 91% if codon redesign is allowed. Experimental validation on a subset of 14 representative designs confirmed the absence of dimer formation, with all assemblies yielding clean bands in agarose gel electrophoresis. Our approach offers a robust and reliable solution for oligonucleotide design, particularly in challenging sequence contexts, and represents a valuable tool for synthetic biology and automated gene assembly workflows.
Although concurrent chemoradiation (CCRT) is the standard of care for locally advanced cervical cancer, oncological outcomes remain suboptimal. Noninvasive monitoring of treatment responses could improve therapeutic management. This study investigated the temporal dynamics of extracellular vesicle (EV)-enriched urinary proteomes in patients with locally advanced cervical cancer undergoing CCRT to identify physiobiological shifts and proteins associated with treatment response and survival. Urine samples from 42 patients were collected longitudinally before, 1 month post-, and 3 months post-CCRT completion (126 samples). Urinary EVs were enriched using strong-anion-exchange magnetic beads. Proteomic profiling was performed using liquid chromatography-tandem mass spectrometry. Differential protein abundance and time-varying Cox regression analyses were used to correlate the proteomic signatures with CCRT response and overall survival. Analysis of longitudinal samples yielded 2,352 quantifiable proteins. Of these, 1,055 exhibited significant temporal shifts. Pathway analysis revealed that tumors and virus-associated proteins peaked at 3 months posttreatment, corresponding to the clinical response assessment window. By 3 months, nonresponders exhibited pronounced suppression of adaptive immune pathways and up-regulation of protein homeostasis and degradation pathways. Time-varying survival analysis identified 8 proteins driven by an altered T-complex ring complex/chaperonin containing T-complex protein 1, ribosomal proteins, and immune-related proteins associated with overall survival. Six of these were expressed in cervical tumor tissues based on the Human Protein Atlas. Overall, EV-enriched urinary proteomics provides a noninvasive liquid biopsy method for monitoring biological dynamics during CCRT. The observed immune suppression in nonresponders and the 8-protein prognostic signature offer preliminary insight into survival patterns and could help guide future studies on adjuvant immune-enhancing approaches.
Human primary amine oxidase (AOC3) is a membrane-anchored protein expressed on the luminal surface of the vascular endothelium at sites of inflammation. It binds via protein-protein interactions to the CE-loops of the C22 domain in Siglec-9 and Siglec-10, but the exact mechanism of interaction is unknown on an atomic level. The AOC3 binding to a radiolabeled Siglec-9 peptide can be used to image inflammation by positron emission tomography (PET). In the present study, a combination of computational and wet lab experiments is used to map the binding of the original and designed shorter Siglec-derived peptides to AOC3, to determine what features in the CE-loop are important for the interaction. To select the peptides for wet lab studies, docking and molecular dynamics simulations were first conducted to predict the binding mode of the peptides in the AOC3 active site channel, and MMGBSA calculations to predict binding energies. Selected peptides were further analyzed through microscale thermophoresis to affirm binding and saturation transfer difference NMR (STD-NMR) to determine the binding epitope to AOC3. Our results unveil that the shorter cyclic Siglec-9 and Siglec-10 peptides derived from the CE-loop bind to AOC3. Our results summarize that a tryptophan and an arginine in both peptides are important for the AOC3 interactions, but their binding mode differs slightly. Our results suggest that the designed shorter Siglec-9 peptide could be advantageous in PET imaging of inflammation and the binding properties of the Siglec-10 peptides hint a biological role for the CE-loop of the C2 domains in Siglec-10.
The rise of rapidly mutating viruses poses growing challenges for drug developers in the fight against antiviral resistance. Viral ion channels generally have low mutation rates and are therefore considered emerging targets for a sustainable solution to the problem of resistance. This study reports the redesign of scaffolds for drug candidates targeting viral ion channels. The redesign was powered by a combination of a computational docking protocol that accounts for water molecules (HydroDock) and a subsequent quantum mechanics-based scoring (QMH-L) of the drug candidates. Extending the antiviral amantadine led to new compounds that better match the alternating hydrophobic and hydrophilic patterns of the inner walls of ion channels—a common feature across many viruses. Simplifying the structure yielded a cyclohexylamine-based minimalist scaffold that demonstrates improved antiviral activity compared to other agents such as amantadine and arterolane. SARS-CoV-2 variants served as test systems in laboratory experiments. The new molecular scaffolds presented here provide a strong foundation for designing potent viral ion channel blockers, helping to fill the drug pipeline.
Liver microsomal metabolic stability is a key determinant of in vivo exposure and an essential filter in lead optimization, yet cross-species prediction remains difficult because of heterogeneous metabolic pathways and limited model interpretability. We propose a cross-species multitask learning framework that integrates complementary molecular modalities-SMILES-derived fingerprints (Morgan and MACCS/RDKit), molecular graphs, and in silico absorption, distribution, metabolism, and excretion (ADME)/physicochemical descriptors-to predict binary microsomal stability (unstable: t 1/2 ≤ 30 min; stable: t 1/2 > 30 min) in human (HLM), rat (RLM), and mouse (MLM) liver microsomes. We curated 18,921 PubChem BioAssay measurements (6,685 HLM; 5,753 RLM; 6,483 MLM). Under stratified 10-fold Bemis-Murcko scaffold cross-validation with ensemble prediction and species-specific thresholds, the model achieved AUROC values of 0.811, 0.806, and 0.794 and AUPR values of 0.854, 0.860, and 0.862 for HLM, RLM, and MLM, respectively, consistently outperforming conventional machine-learning and single-task deep-learning baselines. SHapley Additive exPlanations (SHAP) identified transport/permeability indicators, CYP interaction flags, and the lipophilicity-polarity axis as the features most strongly associated with predicted stability. EdgeSHAPer, a graph neural network explanation method based on SHAP, highlighted stabilizing and destabilizing substructures. Recurrent destabilizing attributions were observed in alkene and allylic/benzylic contexts, whereas amide/carbamate motifs exhibited stabilizing attributions, with nitriles and halogens showing context-dependent effects. Fragment-ADME enrichment analysis characterized associations between local structural motifs and whole-molecule properties including lipophilicity, solubility, and blood-brain barrier permeability. This multi-modal, cross-species framework demonstrates that integrating structural encodings with ADME descriptors enhances both predictive performance and interpretability, yielding hypothesis-generating attributions for structural optimization that warrant prospective experimental validation.
Diffuse large B-cell lymphoma (DLBCL) exhibits substantial biological heterogeneity, leading to pronounced variability in patient response to therapy. Accurate drug response prediction is therefore critical for precision treatment but remains challenging in clinical settings where genomic sequencing, a highly informative modality, is frequently incomplete. Existing methods, often developed from cell-line pharmacogenomic datasets or single-modality data, typically assume fully observed molecular profiles and thus show limited robustness under missing genomic data. To address this limitation, a knowledge-enhanced multimodal framework with genomic reconstruction (KeM-DRP) is proposed for individualized drug response prediction in DLBCL. The framework models the central role of genomics by integrating biological prior knowledge through a gene-pathway-biological process hierarchy, enabling robust representation learning from sparse observations. To compensate for missing genomic measurements, a cross-modal genomic compensation module reconstructs genomically informed latent features from routinely available clinical modalities. Furthermore, a genomics-guided adaptive fusion strategy dynamically integrates heterogeneous modalities conditioned on observed or reconstructed genomic representation. Experiments on a real-world DLBCL cohort demonstrate that KeM-DRP consistently outperforms competitive baselines. The reconstructed genomic representation represents most predictive utility, highlighting the robustness and practical value of the framework under incomplete genomic data.
Trypanosoma cruzi ( T. cruzi ), the etiological agent of Chagas disease, remains a major global health threat with limited therapeutic options. Macrophages are central to host defense against T. cruzi , yet their failure to achieve complete parasite clearance contributes to disease progression. Although 5’ isomiRs have emerged as functional regulators of gene expression, their expression alternations in T. cruzi -infected macrophages remain largely unknown. Here, we performed comparative analyses of 5’ isomiR expression in T. cruzi -infected cells and identified a marked, selective increase in microRNA 5’-end heterogeneity in THP-1-derived macrophages but not in cardiomyocytes or epithelial cells. Comparative 5’ isomiRome analysis revealed 68 differentially expressed 5’ isomiRs in infected macrophages, most of which were specific to T. cruzi relative to other pathogens. Among them, 56 were associated with Argonaute proteins, including 3 derived from the miR-1246 precursor. Focusing on miR-1246|+1, a 5’ isomiR generated by a 1-nucleotide downstream shift at the 5’ end of miR-1246, we showed that its overexpression substantially down-regulated target genes involved in nuclear factor-κB (NF-κB) signaling, circadian rhythm, cytokine responses, and cell migration. Notably, miR-1246|+1 broadly suppressed NF-κB family transcription factors and their downstream effector genes, thereby inhibiting proinflammatory M1 macrophage phenotype. Intriguingly, the suppressed NF-κB regulators Nfkb1 and Rela, together with M1 macrophage signature genes Ccl8, Il1a, Il1b, and Tlr2, displayed rhythmic expression in murine peritoneal macrophages. Collectively, these findings reveal cell type- and pathogen-specific reprogramming of the 5’ isomiR landscape in T. cruzi -infected macrophages and identify miR-1246|+1 as a potential posttranscriptional regulator of macrophage inflammatory polarization.
Experimentally resolved membrane-protein structures have increased substantially, yet annotations remain fragmented across resources differing in scope, curation criteria, and identifier conventions, complicating cross-database comparison and downstream analysis. We present MetaMP, a membrane-protein reconciliation and benchmarking platform that harmonizes metadata from MPstruc, RCSB PDB, OPM, and UniProt into a unified, searchable resource integrating 4,089 unique structures. MetaMP provides provenance-aware discrepancy analysis, quality-control workflows, and 2 assistive modules as reproducible baselines: (a) a broad structural-group classifier trained on OPM-derived membrane-orientation descriptors and (b) a transmembrane-segment benchmarking layer integrating sequence-based and structure-derived topology sources. Cross-source comparison identified 121 broad-group conflicts between MPstruc and OPM (2.96% of 4,089 harmonized entries). These contested cases were expert-reviewed to form a 121-record discrepancy benchmark. Under strict label matching, OPM agreed with expert annotations for 96 of 121 records (79.34%), while the MetaMP assistive classifier agreed for 25 of 121 (20.66%) and MPstruc for 17 of 121 (14.05%). Under benchmark-aware evaluation applying a biologically motivated label-collapsing rule, agreement reached 88.43% for MetaMP, 80.99% for OPM, and 78.51% for MPstruc. In a 24-participant task-oriented user study, structured training was associated with faster task completion, and the adapted SUS-style usability score averaged 72.81, placing the system in the above-average to good range. MetaMP is not intended to replace primary databases, but to make disagreement among them explicit, traceable, and biologically interpretable, providing a reproducible framework for annotation harmonization, expert-guided curation, and membrane-protein benchmarking. Source code and deployment materials are available at https://github.com/Ebenco36/MetaMP-Server .
Aspergillus oryzae (koji mold) is a key microorganism in traditional food fermentations including soy sauce, sake, and miso and is important in novel culinary applications and modern biotechnology, such as sustainable meat alternatives and enzyme production. Despite its industrial importance, until recently, the most recent genome-scale metabolic model (GEM) for A. oryzae dated back to 2008 and was limited to a single strain (RIB40). Here, we present pAo , a pan-GEM for A. oryzae , integrating genomic data from 187 strains to capture species-wide metabolic diversity. Our model comprises 2,025 reactions, representing a 52% increase in metabolic coverage over the RIB40-based model and includes previously overlooked pathways, such as cytochrome P450-mediated xenobiotic metabolism and extended amino acid metabolism. Using this pan-GEM, we derived strain-specific GEMs and validated 8 of them through high-throughput phenotypic screening on 285 substrates. Growth experiments on 4 industrially relevant carbon sources revealed substantial interstrain metabolic diversity, although flux balance analysis indicated that this variability originates at the regulatory rather than network-structural level. This resource provides a foundation for informed strain selection for biotechnological applications and future metabolic engineering in A. oryzae .
Background: Overweight, obesity, and metabolic dysfunction-associated fatty liver disease (MASLD) are increasingly prevalent in adolescents and are linked to alterations in the gut-liver axis. Gut microbiota may contribute to early metabolic disturbances preceding overt disease. Objective: To compare gut microbiota composition and fecal metabolite profiles, including short-chain fatty acids (SCFAs) and amino acids (AAs), between adolescents with overweight/obesity and normal-weight peers, and to evaluate differences associated with hepatic steatosis assessed by FibroScan-controlled attenuation parameter (CAP). Of 111 participants aged 13 to 17 years, 83 and 28 represented the study group with overweight or obesity and normal-weight control group, respectively. Hepatic steatosis consistent with MASLD was defined as CAP ≥250 dB/m. Gut microbiota was assessed by 16S ribosomal RNA (rRNA) sequencing and fecal metabolites by gas chromatography-mass spectrometry (GC-MS). Results: Adolescents with overweight/obesity showed impaired metabolic profiles compared with controls, while overall microbial α and β diversity did not differ between groups. Differences were observed in the relative abundance of several bacterial genera, including depletion of Lactobacillus. Fecal concentrations of most AAs were significantly elevated in the overweight/obesity group, whereas SCFA levels were unchanged. CAP-based stratification revealed that hepatic steatosis was associated with differences in richness-related α-diversity indices without β-diversity separation. Adolescents with fatty liver exhibited a distinct microbial signature involving low-abundance taxa and higher fecal acetate and butyrate levels. Conclusions: Adolescents with overweight/obesity and fatty liver display subtle but biologically relevant alterations in gut microbiota composition and fecal metabolite profiles despite preserved global microbial diversity. Early metabolic impairment appears to be associated with changes in specific bacterial taxa and AA metabolism rather than generalized dysbiosis.
Experimental Objective: Arctic soils harbor diverse bacterial communities that regulate carbon and nutrient cycling under persistent coldness, low-nutrient availability, and episodic freeze-thaw. However, the extent to which bacterial community stability versus spatial specialization structures these communities remains unresolved. Methods: This study characterized bacterial assemblages (microbiota) across 6 geographically distinct Svalbard soil sites using 16S ribosomal RNA gene sequencing, integrated with alpha- and beta-diversity comparisons, taxonomic profiling, linear discriminant analysis effect size (LEfSe) candidate biomarkers, core microbiota analysis, co-occurrence network, and predictive functional inference. Results: Proteobacteria dominated across Svalbard soil sites, accompanied by Actinobacteria, Acidobacteria, and Bacteroidetes. Taxa compositions and beta-diversity revealed spatial structuring, with Wahlbergøya and Northernmost Island relatively more unique community configurations. Consistently, LEfSe identified candidate site-specific taxa, i.e., sulfur-oxidizing bacteria and Acidobacteria, suggesting local environmental selection. The conserved core phylotype analysis indicated microbiota patterns associated with localized geographics and unraveled 48-phylotype core microbiota spanning 5 phyla. Functional predictions demonstrated dual-functional structures: (a) conserved functions in genome maintenance, membrane transport, and central carbon and amino acid metabolisms and (b) site-specific functional enrichments linked to stress responses, redox balance, xenobiotic degradation, and utilization of recalcitrant substrates. Integration of core microbiota taxonomic, network, and functional analyses further revealed 4 mechanistic resilience strategies: genome integrity maintenance, membrane adaptation, oxidative stress alleviation, and flexible resource (i.e., recalcitrant substrates) utilization. These strategies highlighted how shared core taxa collectively translated into microbial adaptive process and persistence, across sites. Conclusions: These findings underpinned that High Arctic soil microbiota were structured consistently with functional plasticity that supported Arctic ecosystem processes.
