Search Results (2,054,932)

Similar to supply chains (SCs), where bullwhip effect shocks spread backward, financial time series exhibit volatility driven by investors’ choices, macroeconomic context, and technological development, among other factors. While deep learning has enhanced the predictive performance for stock prices, limited research has examined the impact of multi-echelon forecasting and shock propagation on predictive performance. This paper presents a reliable predictive framework based on RNNs, transformers, and XGBoost, combined with Automated Machine Learning (AutoML), to design a customizable pipeline for forecasting stock prices and indices throughout the semiconductor SC, from manufacturing and hardware assembly to final demand and carbon emissions. The Hyperparameter Optimization (HPO) used Tree Parzen Estimator (TPE) and BOHB to simultaneously tune hyperparameters in a high-dimensional space. The feature set of each echelon incorporates the previous echelons’ close prices, in addition to the autoregressive components of the current echelon’s target, without leakage. To foster users’ interactivity, the performance analysis gauges the models’ accuracy and complexity per hyperparameter landscape and determines, through SHAP analysis, the hyperparameters and the features driving optimal performance. Moreover, a shock propagation is performed to follow the impact of volatility in upstream variables on forecasts of the final demand and carbon emissions. Full article

Background/Objectives: Pancreatic ductal adenocarcinoma (PDAC) has exceptionally high mortality, largely because of late diagnosis, while established blood biomarkers such as CA19-9 have limited sensitivity for early detection. We evaluated whether modelling serum proteomic measurements as supervised pairwise statistical graphs using synolitic graph neural networks (SGNNs) could provide a useful framework for pre-diagnostic PDAC classification. Methods: We analysed 217 pre-diagnostic serum samples from UKCTOCS, comprising 100 PDAC samples from 75 women and 117 control samples, profiled for 97 proteins. Samples were divided into a development set collected 0–1 year before diagnosis (n = 110) and a temporal holdout set collected 1–2 years before diagnosis (n = 107); 25 PDAC participants contributed samples to both windows. The final SGNN used fold-internal mutual-information selection of 30 proteins, pairwise RBF-SVM-derived edge information, minimum-connected sparsification, and a 10-member GATv2 ensemble per cross-validation fold. Conventional machine-learning comparators were tuned using development data only. Results: The SGNN achieved a mean 5-fold development ROC-AUC of 75.33 ± 10.43%, lower than the tuned Random Forest (81.67%); the other tuned comparators achieved 77.67% (XGBoost), 77.17% (logistic regression), and 73.17% (SVM). Averaging predictions across the five-fold-specific SGNN ensembles yielded a temporal-holdout ROC-AUC of 64.35% (95% CI 52.6–74.6%). In the participant-independent subset of 25 PDAC cases not represented in development and 57 controls, ROC-AUC was 66.04% (95% CI 52.56–78.88%). Conclusions: SGNNs provide a feasible graph-based representation of pre-diagnostic proteomic data but did not outperform optimised conventional machine-learning methods within development cross-validation. The temporal and participant-independent performance estimates warrant further investigation in larger fully independent cohorts. A graph-specific predictive advantage was not directly tested—no such advantage was demonstrated in this dataset—and the clinical utility has not been established. Full article

Metabolic differences are an important factor influencing the quality of chicken meat. To investigate the differences in muscle flavor precursors between Daweishan mini chickens and Tegel broilers, integrated metabolomics analyses were used to compare the muscle metabolite profiles of the two breeds at marketable age (150-day-old Daweishan mini chickens and 45-day-old Tegel broilers). The differential metabolites (DMs) were significantly enriched in aminoacyl-tRNA biosynthesis; glycine, serine and threonine metabolism; glutathione metabolism; alanine, aspartate and glutamate metabolism; arginine biosynthesis; purine metabolism; pantothenate and CoA biosynthesis; cysteine and methionine metabolism; glycerophospholipid metabolism; and D-glutamine and D-glutamate metabolism, which contributed to the quality differences between the two breeds. Moreover, 3-methylhistidine, L-carnosine, L-anserine, DL-lactate, choline, L-phenylalanine, and L-ascorbic acid were present at higher relative abundances in the muscle of Daweishan mini chickens at marketable age. This study provides basic data on the mechanisms underlying the meat quality of local chickens and a reference for the development and utilization of local chicken breeds and the selective breeding of high-quality broilers. Full article

WASHC2C encodes a WASH-complex scaffold implicated in actin remodeling, endosomal organization, and intracellular trafficking, yet its cancer-specific molecular and clinical relevance remains insufficiently defined, particularly in the context of the closely related paralogue WASHC2A. We performed an integrated pan-cancer multi-omics analysis incorporating transcriptomic and protein-level evidence, clinicopathological and survival associations, genomic and epigenetic alterations, pathway activity, gene-set enrichment, immune features, immunotherapy-related observations, and single-cell functional states. Primary RNA analyses extracted both paralogues from the same GDC STAR files across eight cancer types, while covariate-adjusted Cox models evaluated LUAD, BLCA, and LIHC. A locked 50-gene co-expression signature was additionally assessed in independent GEO cohort, GSE31210. The harmonized expression analysis comprised 4365 primary tumors and 429 normal specimens and revealed pronounced lineage-specific variation together with strong WASHC2A–WASHC2C co-expression. Although HNSC showed platform-dependent directional differences between STAR and GSCA, the integrated analyses identified LIHC as the clearest prognostic context: higher WASHC2C expression was associated with poorer overall survival in the primary adjusted model (340 patients; 115 deaths; HR = 1.61 per unit log2[TPM + 1], 95% CI 1.18–2.21; FDR = 0.035), although the magnitude of association varied with model specification. Favorable unadjusted observations in LUAD and BLCA were not retained after covariate adjustment, and primary BLCA OS/DSS models were non-estimable. Across cancers, copy-number variation showed a consistent positive relationship with transcript abundance, whereas pathway, immune, methylation, and single-cell associations demonstrated marked tumor-context specificity. In GSE31210, 48 measurable genes from the locked signature were associated with favorable survival in 204 patients, supporting the relevance of the broader co-expression program, although neither WASHC2C nor WASHC2A contributed directly to that score. Collectively, this study establishes a rigorous multi-omic framework for WASHC2C, highlights its lineage-dependent molecular and prognostic relevance, and prioritizes specific cancer contexts—particularly LIHC—for mechanistic and translational investigation. These findings provide a strong foundation for future paralogue-resolved functional studies and independent clinical validation. Full article

Transjugular intrahepatic portosystemic shunt (TIPS) is an effective treatment for portal hypertension, but post-TIPS hepatic encephalopathy (HE) remains a major complication. Conventional clinical scores use limited clinical and biochemical variables and cannot fully characterize preoperative imaging phenotypes. Although contrast-enhanced abdominal computed tomography(CT) contains morphological, textural, and spatial cues, available datasets are small, CT sequences are redundant, and local details must be modeled together with broader context. We therefore propose Medical Global Zoom Mamba (MGZ Mamba). A medical visual invariance constraint stratifies CT slices by feature density using the nonzero-pixel ratio and gray-level entropy, allowing multiple training instances to be constructed from each patient sequence. A double-nested parallel aggregate attention module combines multi-receptive-field convolution with global channel recalibration, while a feature rearrangement 2D scan module introduces dynamic windows and spatial-neighborhood rearrangement alongside the original two-dimensional selective scan branch. In patient-level five-fold cross-validation on the Post-TIPS Hepatic Encephalopathy Dataset of Shanxi Medical University (HE-SMU), MGZ Mamba achieved 81.65% accuracy, 82.76% recall, 80.53% F1-score, and an ROC-AUC of 0.829. Relative to the strongest baseline for each metric, accuracy, recall, and F1-score increased by 0.41, 1.07, and 0.56 percentage points, respectively, and ROC-AUC increased by 0.007. These findings support MGZ Mamba as an imaging-based approach for preoperative risk stratification and may provide complementary evidence for clinical assessment and treatment planning. Full article

Oral cancer patients suffer pain that is patient-specific and heterogeneous at the site of the tumor. Protease-activated receptor 2 (PAR₂), a G-protein coupled receptor and an oral cancer mediator, is activated by protease cleavage of an N-terminal ligand. Once cleaved, the receptor is retained at the plasma membrane or re-distributed (trafficking) to internal sites followed by downstream protease-specific signaling. The oral cancer tumor microenvironment (TME) is replete with protases that activate PAR₂, potentially evoking PAR₂ trafficking and signaling that is unique to each cancer and reflecting the protease composition of the individual TMEs. To measure PAR₂ cleavage activity in oral cancer samples, an array of overlapping peptides spanning the PAR₂ N-terminus was assembled. In this feasibility study, we assayed the cleavage activity of individual PAR₂-activating proteases and mixtures of proteases modeling the oral cancer TME. Additionally, we confirmed that PAR₂-cleaving proteases are active in oral cancer patient saliva, a sample that can be collected non-invasively to measure PAR₂ cleavage in patients when they are experiencing pain prior to surgery. Further studies should elucidate the utility of the peptide array to measure PAR₂-activating proteolysis in cancer samples. Full article

Mobile health systems have become a central component of the health sector, reshaping how healthcare is accessed, delivered, and experienced, especially in developing countries with limited infrastructure. Mobile health systems are interactive and enable continuous interaction among patients, health professionals, and mobile technology. While these systems improve access to care and patient engagement, they also raise important concerns about how security is handled in highly interactive socio-technological environments. The current state of security assessment for mobile health systems tends to be biased towards technical security controls without addressing the required user interactions during mobile health system development in relation to providing security measures and evidence. This study addresses this gap by proposing a Phase–Role Interaction Security Assessment framework for mobile Health, namely, the PRISM-mH framework. The framework considers three dimensions: interaction-oriented security development lifecycle phases, the roles of the stakeholders involved, and evidence supporting the security requirements. With its help, security can be assessed continuously while taking into consideration the phases and role interactions. This framework is aimed at enhancing the visibility of security during user interactions, while also offering a systematic reference for researchers and practitioners to assess and enhance their security through the mobile health system lifecycle. Considering that PRISM-mH is proposed to address user interaction, technical control, and governance—factors that have not been considered consistently in past mobile health security assessments—this framework opens the door to more comprehensive and evidence-based security assessment, with a goal of enhancing the security position in mobile health development. The author argues that such a framework can help to propose a structure for an agile-driven approach for future studies and provide scoring-based guidance for informing the level of security maturity in mobile health technologies. Full article

The safe integration of unmanned aircraft systems (UASs) into the National Airspace System
(NAS) requires robust Detect-and-Avoid (DAA) capabilities that maintain appropriate
separation from other aircraft. The RTCA DO-365Well Clear thresholds were developed
primarily for en-route operations, raising questions about their suitability for terminal
airspace, where traffic density, maneuvering, and closure-rate characteristics differ substantially.
This paper presents a large-scale evaluation of the RTCA DO-365 Well Clear
thresholds using the MIT Lincoln Laboratory Terminal Encounter Model (LLTEM) V1.0
dataset, which includes one million empirically sampled simulated terminal encounters
generated from a Bayesian model fitted to FAA terminal radar data. Encounter severity was
determined based on the true closest point of approach, against the standard Near Mid-Air
Collision (NMAC) definition. The default en-route Loss ofWell Clear (LoWC) criterion was
then evaluated against those labels and compared jointly with the terminal-area thresholds
proposed by Vincent et al. Scored strictly against NMAC severity, the en-route default
produces a 2.8% false-alarm rate at a precision of 0.27, with no missed NMAC encounters.
Of the three threshold parameters, only the predictive τmod term moves that rate. Vincent
et al.’s full recommended range reduces it further at no detection cost, their lower bound
reaching 0.9% at a precision of 0.54, representing a 67% relative reduction and the best of
the configurations they propose. The predictive term’s constant-velocity extrapolation is
relatively accurate for the duration of most of the encounters in which it engages: 81% of
severity labels are correct at the en-route default, rising to 86% at the tightest configuration tested. These findings provide an evidence-based assessment of the operational suitability
of current Well Clear thresholds for terminal-area UAS operations and support future
refinement of DAA standards. Full article

We profiled 132 RNA-sequencing libraries representing 11 tissues from six Tarim and six Tianshan adult male red deer maintained under common feeding and management. Tissue identity dominated expression structure among 16,308 expressed genes. Across three differential-expression model specifications, 1253 robust group-associated gene–tissue pairs were identified, involving 1184 genes and concentrated in rumen and rectum. Directional enrichment returned 77 terms meeting both within-set and study-wide false-discovery thresholds. Tianshan-group rumen genes were enriched for cytoskeletal, adhesion, ion-transport and circadian annotations, whereas Tarim-group reticulum genes were enriched for cell-surface and immune-signaling annotations. Cross-tissue recurrence was uncommon and identified a small set of annotated genes and provisional loci for follow-up. Complementary whole-genome principal component analysis (PCA) separated the recorded groups across two genotype-depth filters; weighted Weir–Cockerham F_ST was 0.341 and 0.325. These results describe tissue-resolved expression patterns and genetic structure in the sampled cohort. Functional, taxonomic and population-wide interpretations require independent validation. Full article

Cisplatin is a highly effective anti-neoplastic agent, but its clinical application is limited by its association with severe nephrotoxicity. Although extensive research has identified several mechanisms underlying cisplatin-induced nephrotoxicity, the precise molecular pathways remain incompletely understood. Recent studies have highlighted the role of intrinsic survival, senescence, and longevity signaling pathways, including those governed by the FOXO3 transcription factor. In this study, we examined the effects of cisplatin and curcumin on the DNA methylation status and mRNA expression of FOXO3 in human embryonic kidney (HEK293) cells. Our results indicate that the FOXO3 promoter is methylated in cisplatin-treated cells, and that cisplatin regulates FOXO3 transcription in a dose-dependent manner. Quantitative RT-qPCR revealed that moderate cytotoxic concentrations of cisplatin (10 and 40 µM) significantly downregulate FOXO3 mRNA. Furthermore, co-treatment with the antioxidant curcumin successfully restored and enhanced FOXO3 expression at moderate cisplatin doses (10 µM) but failed to rescue expression at high doses (40 µM). Moreover, curcumin treatment (10 µM) reverses the effects of low-dose, but not high-dose, cisplatin (40 and 60 µM). Taken together, these findings provide a foundation for understanding the epigenetic and transcriptional regulation of FOXO3 and highlight its potential as a target for antioxidant-mediated nephroprotection and its role in cell survival. Full article

Unattended photovoltaic monitoring nodes must maintain authenticated operation under physical access, intermittent connectivity, and environmentally induced key-reconstruction errors. We present a distributed adaptive hardware security framework (DAHSF) combining trusted-gateway threshold authorization, adaptive static random-access memory (SRAM) physical unclonable function (PUF) reconstruction, and incident-driven inspection. Independent epoch secrets separate current authorization from retained historical shares. A nested Bose–Chaudhuri–Hocquenghem (BCH) error-correction scheme reconstructs a fixed 2016-bit SRAM-PUF response, from which universal hashing extracts a 256-bit seed; federated estimation guides decoding effort. The cost and defense trade-offs are assessed under the tested workload, gateway trust, prediction-calibration, and availability conditions. On a 30-node ESP32 platform, DAHSF reduces median startup and mean node power by 4.8% and 4.4% relative to a matched lightweight baseline, with 14.3% more incremental RAM. Predicted-bound decoding reduces mean decoding time from 96 to 59 microseconds. A separate confirmatory cohort records three failures in 120,000 stressed reconstructions; its board-clustered 95% upper bound is $8.6\times {10}^{-5}$. In the moderate attack scenario, the full inspection profile reduces compromised-node exposure by 30.0% relative to a risk-threshold policy and by 18.0% relative to frozen propagation-rate estimates. The recorded control workload has no observed 1 ms MPPT deadline misses, while fault injection still bypasses authorization in 168 of 4000 attempts. Full article

Land surface temperature (LST) is closely linked to land–atmosphere energy exchange, while conventional point forecasting cannot quantify forecast uncertainty or explicitly enforce physical consistency over long horizons. We propose Physics-former-Rn-SG to address two limitations: fixed physical penalties cannot adapt to state-dependent forecast reliability, while directly coupling uncertainty to physics weighting can encourage variance inflation. The model integrates a variable-wise iTransformer, NIG evidential prediction, supervised future -forcing heads, and a reduced Rn-SEB residual. A stop-gradient, normalized, and clipped weighting scheme enables uncertainty-adaptive physical regularization while preventing the physics loss from directly driving variance inflation. Evaluated on MPI-Saale and DE-Gri data from 2024 to 2025, the model achieved the lowest mean MAE in all 16 site horizon combinations across 96–384 h forecasts. Its average absolute MAE advantage was 0.157 °C over NIG-iTransformer and 0.046 °C over fixed Rn-SEB, with the latter improvement smaller and more variable across sites, horizons, and validation protocols. Multi-level coverage error, CRPS, WIS, NLL, and ECE were also lowest. The variance inflation ratio decreased from 1.323 to 1.019 with stop-gradient. In the matched DE-Gri full-flux benchmark, reduced Rn-SEB maintained prediction accuracy while reducing training and inference time by about 11.0% and 6.5%, respectively. Overall, Physics-former-Rn-SG balances forecast accuracy, calibration, and physical consistency, with potential as a station-scale forecasting or post-processing tool for uncertainty-aware environmental applications. Full article