Browse Articles

Discover research articles across all indexed journals

Optimization of low-carbon multi-temperature joint distribution for fresh agricultural products under 3D loading constraints

PLoS ONE Juping Shao, Fan Gao, Yanan Sun Jul 13, 2026 DOI: 10.1371/journal.pone.0353789

With the growing demand for fresh agricultural products, improving the efficiency and sustainability of cold-chain distribution has become increasingly important. Multi-temperature joint distribution provides an effective solution for serving products with different temperature requirements, yet its implementation remains challenging due to the need to coordinate vehicle routing, three-dimensional loading, and carbon-emission reduction objectives. To address this issue, this paper develops a low-carbon multi-temperature joint distribution optimization model under three-dimensional loading constraints. A hybrid algorithm integrating genetic algorithm and tabu search is proposed to solve the model efficiently. The proposed approach is validated using real-world data collected from a fresh agricultural products supply chain company. The results show that the multi-temperature joint distribution mode reduces total operating costs by 30.04% and carbon emissions by 30.62% compared with the conventional single-temperature distribution mode. Moreover, the proposed hybrid algorithm achieves faster convergence and better solution quality than the conventional genetic algorithm. These findings demonstrate the effectiveness of integrating three-dimensional loading, multi-temperature distribution, and low-carbon objectives within a unified optimization framework, providing practical support for distribution planning and decision-making in cold-chain logistics.

Corrigendum: Enhancing antiPD‐1 Immunotherapy by Nanomicelles Self‐Assembled from Multivalent Aptamer Drug Conjugates

Angewandte Chemie International Edition Jul 13, 2026 DOI: 10.1002/anie.9343292

Ultrafast and Persistent Electrolyte Acid Scavenger for Stable High‐Voltage Lithium Batteries

Angewandte Chemie International Edition Jiachao Duan, Guohuang Kang, Qi Liu et al. Jul 13, 2026 DOI: 10.1002/anie.8016076

ABSTRACT Hydrofluoric acid (HF) generation in carbonate‐based electrolytes with LiPF 6 is a primary failure cause in high‐voltage lithium‐ion batteries. Here, a new dual‐functional additive 5‐trimethylsilanylthiazole (5‐SiTZ) is developed by rational molecular design for ultrafast and persistent acid scavenging. Siloxane group in 5‐SiTZ reacts spontaneously with HF, while its thiazolyl moiety acts as a secondary Lewis base scavenger and chelates PF 5 to suppress further HF generation. This synergistic action achieves ultrafast (<1 s) and complete HF removal for over 200 days upon open‐air exposure. Furthermore, 5‐SiTZ promotes the formation of a robust, inorganic‐rich cathode electrolyte interphase. The combined effect of persistent HF removal and interface stabilization enables commercial graphite||NCM811 pouch cells to deliver outstanding capacity retention of 92.1% after 700 cycles at 4.6 V. This work establishes a paradigm of synergistic molecular design for multi‐functional additives, providing a practical route to durable high‐energy batteries.

Crystallographic registry in epitaxial V3O5 thin films

Applied Physics Letters Armando Rúa, Alexander Bartenev, Angel D. Reyes-Diaz et al. Jul 13, 2026 DOI: 10.1063/5.0336965

Crystallographic orientation plays a central role in determining transport properties in correlated oxides with low-hysteresis metal–insulator transitions. Here, V3O5 thin films were grown on Al2O3 (0001) and their crystallographic registry was evaluated using x-ray diffraction and cross-sectional transmission electron microscopy. θ–2θ measurements show a single out-of-plane orientation defined by the (202) reflection. Azimuthal φ-scans of the V3O5 (310) reflection exhibit 12 discrete maxima over 360°, indicating a finite set of in-plane orientations imposed by the substrate symmetry. High-resolution TEM and fast Fourier transform analysis confirm crystallographic coherence at the film–substrate interface. The electrical conductivity exhibits a metal–insulator transition at TMIT ≈ 423 K with no measurable thermal hysteresis and a total change of ∼1.6 orders of magnitude between 300 and 480 K. These results establish crystallographic registry in epitaxial V3O5 thin films and provide a structurally well-defined system for future studies of orientation-dependent transport.

Comparative study of different methods for bathymetric mapping based on multi-source remote sensing data

Scientific Reports Tianqi Lu, Luyi Wang, Caiying Ji et al. Jul 13, 2026 DOI: 10.1038/s41598-026-61957-x

Comparison of Moducare versus Wait-and-See approach for histologically proven Low-grade Cervical Intraepithelial Neoplasia (CIN1) (MODUCIN1 TRIAL) – Study protocol

PLoS ONE Dimitrios Zouzoulas, Iliana Sofianou, Kimon Chatzistamatiou et al. Jul 13, 2026 DOI: 10.1371/journal.pone.0353119

Human Papillomavirus (HPV) is causally associated with cervical cancer and precancerous lesions (dysplasias) of the cervix. The treatment of choice for low-grade lesions is monitoring with no-treatment, because the majority of them will regress spontaneously. However, this wait-and-see approach can be assisted by nutritional supplements for immune support, like Moducare. The hypothesis of the trial is that Moducare can enhance the natural regression of CIN1. MODUCIN1 is a prospective, open-label, randomized trial where eligible patients will be randomized (1:1) to either the wait-and-see approach (control arm) or to the six months oral administration of Moducare capsules. The main inclusion criterion is newly histologically proven CIN1 and the main exclusion criterion is any previous cervical intraepithelial neoplasia with or without treatment or history of pelvic malignancy. The primary objective is to compare the regression rates of low-grade cervical intraepithelial neoplasia (CIN1) between the two groups: Wait-and-see approach and Moducare. The sample size is estimated at 182 eligible patients, while accrual is expected to last one year. The primary endpoint is expected to be reached after six months from last patient enrollment. The Trial Registration Number is NCT07379905.

Beyond <i>para</i> ‐ and <i>meta</i> ‐Substitution: Synthesis and Biological Validation of Chalcogen‐Rich Thiabicyclo[3.1.1]heptanes as Bioisosteres of Aryl Thioether Derivatives

Angewandte Chemie International Edition Kun‐Ju Wang, Lei Tang, Heng‐Xian He et al. Jul 13, 2026 DOI: 10.1002/anie.5228347

ABSTRACT While bicyclo[3.1.1]heptanes (BCHeps) and oxa‐/aza‐BCHeps have emerged as valuable arene bioisosteres, the corresponding thia‐analogs remain underdeveloped. Notably, synthetic efforts to access aryl thioether mimetics have focused predominantly on strain‐release thiofunctionalization of [1.1.1]propellane or [3.1.1]propellane to access para ‐ and meta ‐substituted surrogates, whereas three‐dimensional analogs of ortho ‐ and 1,2,4‐trisubstituted aryl thioethers remain elusive. Herein, we report a one‐pot stepwise protocol featuring anti ‐thio(seleno)sulfonylation/annulation of bicyclo[1.1.0]butanes (BCBs) to enable scalable access to functionalized 2‐thiabicyclo[3.1.1]heptanes (thia‐BCHeps) under mild conditions. The synthetic utility is further demonstrated by diverse derivatization of these thia‐BCHeps building blocks and facile access to bioisosteres of ortho ‐, meta ‐, and 1,2,4‐trisubstituted aryl thioether derivatives (e.g., aryl sulfones and aryl sulfoximines). DFT calculations revealed that the reaction proceeded via a polar addition pathway, involving nucleophilic attack of cesium methanesulfinate with BCB. The observed diastereoselectivity originates from the stabilization of the anti ‐addition transition state by a favorable π–π stacking interaction between two phenyl rings. Crystallographic analysis revealed that these thia‐BCHeps display geometric properties almost identical to those of ortho‐ , meta‐ , and 1,2,4‐trisubstituted aryl thioethers. Physicochemical studies and biological evaluation further validated these thia‐BCHeps as a new generation of saturated bioisosteres for aryl thioethers.

Synergistic dual built-in electric fields enabled self-powered humidity sensors based on MXene/MoS2 van der Waals heterostructures

Applied Physics Letters Xing Li, Mingshun Qi, Yongpeng Wu et al. Jul 13, 2026 DOI: 10.1063/5.0334119

Developing zero-power humidity sensors is critical to overcoming battery limitations in wearable health monitoring. Here, we report a high-performance self-powered humidity sensor based on MXene/MoS2 van der Waals heterostructures. A thickness gradient in the chemical vapor deposition-grown MoS2 film creates a stable in-plane built-in electric field, while the work function difference between MXene and MoS2 forms a Schottky junction that suppresses dark current and boosts the signal-to-noise ratio. In a humid environment, the super-hydrophilicity of MXene induces moisture-mediated charge transfer and electron injection. The vertical interfacial field efficiently pumps electrons into the MoS2 channel, where they subsequently undergo directional drift driven by the in-plane built-in field, generating a stable direct current output. The sensor exhibits excellent linearity (R2 &amp;gt; 0.99) and ultrafast response/recovery. It enables high-sensitivity monitoring of human respiration and voiceprint characteristics. This dual-field synergistic mechanism offers a promising strategy for next-generation low-power, high-sensitivity environmental and health-monitoring microsystems.

The oral microbiome is associated with stress, adversity, and mental health in young adults

Scientific Reports Naomi N. Gancz, Paul W. Savoca, Bridget L. Callaghan Jul 13, 2026 DOI: 10.1038/s41598-026-60803-4

Abstract Acute stress responsivity and early-life adversity are both associated with oral microbiome differences that lead to increased mental and physical health risk. However, the joint association of stress responsivity and early adversity with the oral microbiome is not yet understood. Additionally, while the link between the oral microbiome and cortisol has been investigated, another major component of stress responsivity – the parasympathetic response – has been relatively neglected. Therefore, we examined parasympathetic functioning during an acute stressor, retrospective early adversity, oral microbiome (using the bacterial 16S gene), and anxiety and depression symptoms in 76 undergraduates. Early adversity was associated with lower microbiome diversity. Independently of adversity, parasympathetic withdrawal during the stressor was positively associated with Alysiella , a genus previously linked to latent viral infections. Early adversity and parasympathetic function did not significantly interact. Additionally, depressive symptoms were negatively associated with Butyrivibrio , which produces neuroprotective metabolites. Our findings suggest that childhood adversity and parasympathetic stress response are independently associated with oral microbiome differences, and that the oral microbiome is associated with mental health. By characterizing the associations of the oral microbiome with stress, adversity, and mental health, this study lays important groundwork for future research on psychophysiological causes and outcomes of oral microbiome differences.

Evolving optimal text clusters: A novel GA-driven framework for dynamic ensemble fusion of multi-model contextual embeddings

PLoS ONE Ali Sabah, Zaid Alaa Jul 13, 2026 DOI: 10.1371/journal.pone.0353506

Text clustering is an essential activity in unsupervised natural language processing (NLP), and allows the automatic structure of large-scale textual collections (in natural language processing) in news classification, routing of technical questions, and text summarisation. With the emergence of contextual embedding models, such as SBERT, RoBERTa, and DistilBERT, there has been a significant improvement in the quality of clustering, with these models producing dynamic and context-sensitive representations that are more likely to reflect domain-specific semantics than the traditional word embeddings of Word2Vec and GloVe. Although the models of contextual embedding have their own advantages, they show varying performance in different domains and the current ensemble techniques propose fixed, flat fusion weights which do not utilize their complementary abilities. There is no current solution to dynamic, label-free optimisation of weight to multi-model contextual embedding fusion in unsupervised clustering which would bridge a major gap between fixed integrative approaches and adaptive, domain-accommodative model combination. This paper presents a new Genetic Algorithm (GA)-based ensemble model that dynamically optimizes the best fusion weights of SBERT, RoBERTa, and DistilBERT embeddings without ground-truth labels. The hypothesis is that evolutionary optimisation can discover domain-adaptive weight sets that outperform standalone models as well as fixed ensemble baselines on linguistically heterogeneous datasets. The framework uses L2-normalisation and topology in which the topology is a sum-to-one constraint so that the models can be fairly integrated, and a composite fitness measure based on Silhouette Score, Adjusted Rand Index, and Topic Coherence to direct the weight evolution through tournament selection, uniform crossover, and Gaussian mutation. The proposed framework on three heterogeneous benchmarks, AG News, 20 Newsgroups, and Stack Overflow, has Silhouette Score improvements of +14–16 percent, and Topic Coherence gains of +17 percent, all statistically significant at p &lt; 0.05. Evolved weights can be completely understood, depicting domain-specific model roles, which guide future architecture choice. The framework is suggested as a scalable, modular system to practical unsupervised NLP problems, and its architecture can be easily extended to new transformer systems and semi-supervised systems.

Beat-frequency sub-terahertz generation via photomixing using dual continuous-wave multimode laser diodes

Applied Physics Letters Yuanhao Zeng, Verdad C. Agulto, Kosaku Kato et al. Jul 13, 2026 DOI: 10.1063/5.0344954

Terahertz time-domain spectroscopy (THz-TDS) is an advanced technique for material evaluation. TDS commonly uses femtosecond pulsed lasers as the light source. Recent studies have shown that continuous-wave multimode laser diodes (CW-MLD) can be an effective complementary choice. However, CW-MLD-TDS has a discrete spectral shape with fixed frequency peaks. In this research we constructed a THz-TDS driven by dual MLDs and achieved control of internal longitudinal mode frequencies by regulating MLD temperatures. This enabled the generation and tuning of new components corresponding to the beat frequencies of longitudinal modes between MLDs. This advancement will enhance the practicality of MLD-THz-TDS for spectroscopic analysis.

Restriction spectrum imaging reveals brain microstructural correlates of obesity risk in early childhood

Scientific Reports Mohammadreza Bayat, Bianca Braun, Madeline Curzon et al. Jul 13, 2026 DOI: 10.1038/s41598-026-61029-0

Transcriptomic characterization of key psoriasis-associated genes based on single-cell RNA-seq and machine learning

PLoS ONE Weixiang Wang, Qiang Zhang, Suting Xu Jul 13, 2026 DOI: 10.1371/journal.pone.0352663

Background Psoriasis is a multifaceted skin and systemic disorder driven by a complex interplay of genetic, immunological, and environmental factors. Genetic predisposition plays a pivotal role, with the IL-17/IL-23 immune axis recognized as a central pathogenic pathway. Ongoing research, however, continues to uncover additional critical drivers, cytokines, intracellular signaling networks, and potential therapeutic targets. Methods Single-cell RNA sequencing (scRNA-seq) datasets comprising both psoriatic and healthy samples were obtained from the Gene Expression Omnibus (GEO). Cell-type proportions were estimated using Cell-type Identification by Estimating Relative Subsets of RNA Transcripts (CIBERSORT), and weighted gene co-expression network analysis (WGCNA) was applied to explore correlations between cell types and gene signatures. Machine learning algorithms were subsequently employed to identify four psoriasis-associated key genes: DEFB4A , GJB2 , SERPINB3 , and SERPINB13 . Their expression was validated in bulk RNA-seq datasets. Using scRNA-seq data, we further investigated the lesional regulatory roles of these genes and their associated pathway alterations, and we proposed targeted therapeutic strategies. Results A series of algorithms identified 271 hub genes significantly associated with psoriasis lesions and basal cells. Machine learning analysis refined this set to four key genes in psoriasis: DEFB4A , GJB2 , SERPINB3, and SERPINB13 . Conclusions These four psoriasis-associated driver genes were upregulated in lesional skin. We also screened small-molecule compounds targeting these genes, offering potential therapeutic strategies.

Laser direct writing assisted fabrication of CsPbBr3/graphene/Ag nanocomposites

Applied Physics Letters Xiaochen Fang, Guixiang Ding, Xiaotong Zhang et al. Jul 13, 2026 DOI: 10.1063/5.0335785

Composite materials of inorganic perovskites and graphene have proven to be an effective strategy for achieving excellent optical and electrical properties. Building upon this foundation, we introduce noble metals (Ag) to further enhance the optical and electrical performance. By integrating laser direct writing and photo-induced synthesis processes, we successfully fabricated CsPbBr3/graphene/Ag composite materials. This strategy enables the in situ anchoring of Ag nanoparticles and CsPbBr3 nanocrystals (NCs) within a porous graphene framework. Photoluminescence (PL) spectroscopy indicates that graphene can significantly quench the luminescence of pristine CsPbBr3 NCs in CsPbBr3/graphene composites, which may be induced by the rapid separation and transfer of photogenerated hole–electron pairs. Upon the addition of silver, the PL intensity slightly increases, possibly due to the surface plasmon resonance of Ag. Furthermore, the ternary architecture exhibits an enhanced photo-response, with the photocurrent density of the CsPbBr3/graphene/Ag composites reaching approximately twice that of the pristine CsPbBr3 NCs. Reliable photoelectrochemical switching functionality is demonstrated under intermittent visible-light irradiation. Within this ternary network, CsPbBr3 provides active catalytic sites, the graphene framework facilitates rapid electron transport, and the Ag nanoparticles enhance interfacial charge-transfer kinetics. Consequently, these architectures offer promising avenues for advanced photocatalytic applications.

Processing-dependent reinforcing behavior of garlic peel and pineapple leaf fibers in poly(butylene succinate) all-biocomposites

Scientific Reports Supapit Patomnupong, Taweechai Amornsakchai, Ittipol Jangchud Jul 13, 2026 DOI: 10.1038/s41598-026-62423-4

The effects of continuous chest compression protocol on cardiopulmonary resuscitation (CPR) quality and fatigue

PLoS ONE Katie J. Lyman, Thomas A. Hanson, Jennifer A. Longo Jul 13, 2026 DOI: 10.1371/journal.pone.0353187

Aim This study investigated how fatigue (measured both subjectively and objectively) and cardiopulmonary resuscitation (CPR) quality are impacted by various CPR protocols, principally comparing continuous chest compressions with intermittent rest periods. Methods A sample of 23 participants (M = 22.3 years ± 3.8) was used in a cross-sectional design. Four compression:rest ratios were analyzed: 30:10, 50:10, 100:10, and continuous compressions (CC), with all participants performing CPR for 9 minutes in each condition. Subjective fatigue was measured using the Borg Rating of Perceived Exertion (RPE) at baseline, three, six, and nine minutes. Objective fatigue data were collected using the Parvo metabolic cart and heart rate (HR) monitors. Regression models of CPR performance quality (i.e., depth, recoil, and rate) investigated the effect of CPR protocol while controlling for participant variables. Fatigue (i.e., RPE, HR, and VO 2 ) were compared across CPR protocols using repeated measure one-way ANOVA and post hoc Tukey’s honestly significant difference (HSD) tests. Results Significant differences in performance were observed between CC and other protocols: 30 (p &lt; .001), 50 (p = .009), and 100 (p = .031). Percent of full chest recoil was also significantly lower during CC than in 30, 50, and 100 conditions (all p &lt; .001). Subjective fatigue increased significantly with longer compression ratios (p &lt; .001). Conclusion Higher compression ratios result in greater perceived fatigue and reduced CPR performance quality. Rescuer fatigue should be considered in CPR protocol design to optimize both rescuer effectiveness and patient outcomes.

Revealing Spatially Resolved Exciton Dynamics of Dual Active Centers for Dehydrogenation Oxidation of Benzyl Alcohol Coupled by Hydrogen Evolution

Angewandte Chemie International Edition Wen Duan, Wei Li, Rongxia Cui et al. Jul 13, 2026 DOI: 10.1002/anie.4629546

ABSTRACT Photooxidation of biomass alcohols coupling H 2 evolution has received increasing attention due to its higher economy. However, the insufficient photoexciton utilization has become a bottleneck restricting its practical application, thereby precisely regulating the microstructure of photocatalysts is crucial for achieving efficient photoactivity. Strategically introducing functional active centers has been proven to be a valid strategy for optimizing the exciton dynamics of photocatalysts. Herein, the cobalt (Co) clusters and single‐atom Co─O sites were co‐decorated on the heterostructured MoS 2 /CdS composite to integrate the merits of heterojunction and non‐precious metal cocatalysis, which was demonstrated by the x‐ray absorption fine structure spectroscopy and spherical aberration‐corrected transmission electron microscope techniques. Remarkable collaborative photoactivity for benzyl alcohol (BA) oxidation (181.8 mmol BAD ·g −1 ·h −1 ) coupling H 2 evolution (64.6 mmol H2 ·g −1 ·h −1 ) was achieved under artificial sunlight. Femtosecond transient absorption spectroscopy and Kelvin probe force microscopy techniques proved the improved spatial separation/transfer kinetics and hindered recombination kinetics of photoexcitons. Based on density functional theory, the Co clusters act as the oxidation sites for BA, and the Co─O sites serve as the H 2 evolution reduction centers, enabling the synergistically enhanced REDOX half‐reactions due to more effective spatial separation of photoexcitons. This study proposes a strategic catalyst construction insight for highly efficient collaborative REDOX photocatalysis.

Impact of V-defects on current flow and recombination pathways in long-wavelength GaN-based light-emitting diodes

Applied Physics Letters Tsung-Yin Tsai, Tanay Tak, Yuh-Renn Wu et al. Jul 13, 2026 DOI: 10.1063/5.0340858

In long-wavelength GaN-based light-emitting diodes (LEDs), {101¯1} V-defect sidewalls provide a current pathway for lateral electron and hole injection into the (0001) quantum wells (QWs). However, experimental techniques often struggle to distinguish the specific nature of this injection, particularly in determining whether the lateral current flow is governed by drift or diffusion. Moreover, the impact of Shockley–Read–Hall (SRH) recombination within V-defect sidewall (QWs) is largely unexplored. In this work, a two-dimensional self-consistent Poisson and drift–diffusion solver, integrated with localization landscape theory and random alloy disorder in the InGaN QWs to account for carrier transport and quantum effects, is used to study the carrier transport and recombination in and around V-defects. The results demonstrate that the current flows into the planar (0001) QWs from the V-defect rims (the interface region between c-planes and V-defect sidewalls) largely through diffusive transport, indicating that the injection pathways for both electrons and holes are direct and rapid. Compared to the radiative recombination rate in c-plane QWs, the SRH recombination rate within V-defect sidewall QWs is relatively low. This suggests that SRH recombination in V-defect sidewall QWs is a minor factor in long-wavelength GaN-based LEDs. Overall, this study offers mechanistic insights into current flow, revealing how it governs the carrier transport and recombination in V-defect sidewall QWs.

Vector OBF detection method of magnetic tensor for small magnetic targets

Scientific Reports Lei Xu, Chunhang Chen, Ning Zhang et al. Jul 13, 2026 DOI: 10.1038/s41598-026-62120-2

Abstract To improve the detection probability of small magnetic targets and enrich the direction information in magnetic scalar orthonormal basis function (OBF) detection, this paper proposes a vector OBF detection method using the magnetic tensor. Firstly, the OBF expression of the magnetic tensor is determined by analyzing the OBF decomposition principle of the magnetic tensor vector. Then, the magnetic tensor vector OBF detection model is constructed, and the magnetic tensor vector OBF detection method is proposed. Finally, the AR whitening filter is investigated to optimize the proposed method. The detection method can detect small magnetic targets under the condition of a low signal-to-noise ratio (SNR). The simulation results indicate that the detection accuracy of the scalar OBF detection method is significantly decreased under the Gaussian noise with an SNR of −19 dB, while the vector OBF detection method can still effectively detect magnetic anomaly signals; when detecting the same target under the false alarm probability of 1.5% and different fractal noises, the detection probability of the vector OBF detection method is 6 ~ 35% higher than that of the scalar OBF detection method.

Tool choice matters: Evaluating edgeR vs. DESeq2 for sensitivity, robustness, and cross-study performance

PLoS ONE Mostafa Rezapour Jul 13, 2026 DOI: 10.1371/journal.pone.0353788

Differential gene expression (DGE) analysis is foundational to transcriptomic research, yet tool selection can substantially influence results. This study compares two widely used DGE tools, edgeR and DESeq2 , using real and semi-simulated bulk RNA-Seq data sets, mostly from human patients, spanning viral infection, bacterial infection, and fibrotic conditions. We evaluated tool performance across four dimensions: (1) sensitivity to sample size and robustness to outliers; (2) classification performance of uniquely identified gene sets within the discovery dataset; (3) pathway-level concordance of significant DEG sets; and (4) generalizability of tool-specific gene sets across independent studies. First, using Bonferroni-adjusted p -value &lt; 0.05 and absolute log 2 fold change greater than 1 (i.e., | log 2 FC | &gt; 1 ) as significance criteria, repeated subsampling showed that DESeq2 generally identified more Differentially Expressed Genes (DEGs) than edgeR at smaller sample sizes, while the tools became more concordant as sample size increased. Both tools showed similar responses to simulated outliers, with Jaccard similarity decreasing as more swapped samples were introduced. Second, classification models trained on tool-specific genes showed that edgeR achieved higher F1 scores in 9 of 13 contrasts and more frequently reached perfect or near-perfect precision. Third, Hallmark and KEGG pathway enrichment analyses showed that many contrasts retained substantial pathway-level agreement between tools, although selected contrasts still showed tool-specific enriched pathways. Finally, in cross-study validation using four independent SARS-CoV-2 datasets, edgeR -specific genes yielded higher AUC, precision, and recall in held-out datasets, with some test cases achieving perfect separation. Overall, our findings show that DESeq2 may identify more DEGs under stringent thresholds, whereas edgeR often yields more conservative, predictive, and generalizable gene sets. These findings emphasize that DGE tool choice should be guided not only by DEG yield, but also by the downstream reproducibility, predictive value, and biological interpretability of the resulting gene sets.