Search Results (3,071)

The Gravity Recovery and Climate Experiment (GRACE) mascon data relies on minor gravitational field variations to map terrestrial water storage anomaly (TWSA). However, the coarse spatial resolution of three degrees by three degrees restricts their application for evaluating small-scale changes in water storage. To address this challenge, in this study, GRACE and GRACE Follow-On (GRACE-FO) data from 2003 to 2023 were downscaled to 800-m resolution across the Coastal Lowland Aquifer System (CLAS) in Texas, Louisiana, Mississippi, Alabama, and Florida. This downscaling used machine learning (ML) models, including Random Forest (RF), Artificial Neural Network (ANN), and Deep Neural Network (DNN). These models incorporated variables such as anomalies in total precipitation (APT), mean temperature (ATM), normalized difference vegetation index (ANDVI), evapotranspiration (AET) from 2003 to 2023, Shuttle Radar Topography Mission DEM, slope angle, soil type, and lithology to generate monthly 800-m TWSA maps. The ANN model showed strong predictive performance (R² = 0.869–0.989 with low RMSE), although the DNN achieved slightly better statistical accuracy and spatial evaluation metrics; however, ANN was selected for its more realistic and spatially consistent outputs regionally. Building on this improved spatial resolution, analysis of the downscaled TWSA data from 2003 to 2023 identified an overall declining trend in water storage. Trend analysis using linear regression shows that the western CLAS—particularly the Gulf Coast aquifer in Texas and western Louisiana—experiences the strongest depletion, with rates of −0.30 and −0.17 cm/year in Zones 1 and 2, respectively, with Zone 1 being statistically significant. In contrast, the eastern CLAS shows relatively stable conditions, with weak, non-significant increases (+0.05 to +0.18 cm/year), likely reflecting natural variability rather than sustained long-term gain. Therefore, ML-based downscaling of GRACE data enables high-resolution TWS assessment and provides a framework for future extraction of groundwater storage anomalies (GWSA), supporting improved groundwater management. Full article

Chronic inflammation is a well-established risk factor for cancer progression. This review aims to determine how persistent inflammatory signaling reshapes the tissue microenvironment to favor tumor cell proliferation, survival, and progression. It also discusses the role of cytokines such as IL-6 and TGF-β, reactive oxygen species (ROS), and the transcription factors NF-κB and STAT3 in inflammation and in the tumor microenvironment. Sustained activation of these pathways promotes genomic instability, loss of tumor suppressor gene function, enhanced oncogene expression, and resistance to apoptosis, collectively facilitating malignant transformation and tumor development. The key novelty of this review lies in integrating these interconnected networks with new evidence to clarify how they drive cancer initiation and progression. Furthermore, we discuss the therapeutic potential of plant-derived bioactive compounds, with a particular emphasis on curcumin. Curcumin exhibits significant anti-inflammatory and anticancer effects through inhibition of NF-κB and STAT3 signaling and its downstream targets, thereby attenuating inflammation-driven tumorigenesis. However, its clinical application is limited by poor bioavailability. Finally, this review highlights current strategies to overcome these limitations and future directions for optimizing curcumin-based interventions in inflammation-associated diseases. Full article

Background. Preoperative risk stratification of pancreatic neuroendocrine tumors (PNETs) is constrained by the unavailability of histologic grade before resection. We hypothesized that a panel of biologically informed CT-radiomic signatures, combined with patient-level Δ-radiomics referenced to the contralateral pancreas, would support preoperative discrimination of progression and grade in a two-center pilot cohort. Methods. Forty-four patients with histologically confirmed PNET who underwent contrast-enhanced preoperative CT and surgical resection at two academic centers were analyzed. Lesion and contralateral non-tumor-bearing pancreatic parenchyma regions of interest were revised in 3D Slicer by a board-certified pancreatic surgeon and verified intraoperatively against surgical pathology. PyRadiomics v3.0 features were extracted with IBSI-concordant settings. Parametric ComBat batch correction was applied across the two centers (biological-covariate balance verified beforehand), and Δ-radiomic features (lesion combat–pancreas combat) were computed for the 106 intensity/texture primitives. We constructed a panel of biology-informed hybrid signatures partitioned into a preoperative lesion-only family (Family A; seven signatures) and a preoperative Δ-radiomic family (Family B; three signatures). Candidate features were filtered through correlation clustering, baseline-adjusted likelihood-ratio testing with Benjamini–Hochberg FDR control, and 100-bootstrap stability selection. Three predictor blocks were compared per target with three classifiers each (Logistic Regression, Random Forest, Gradient Boosting): M0 (five-variable clinical baseline), MA (M0 + Family A), and MB (M0 + Family B). Discrimination was reported as AUC with bootstrap 95% CI; calibration was assessed using the Brier score and TRIPOD-recommended calibration intercept and slope; and cross-center generalization was evaluated with leave-one-center-out (LOCO) cross-validation. Univariable Cox regression with bootstrap and permutation inference was used for progression-free survival (PFS). Results. The cohort had 16 progression events and eight deaths (median follow-up was 38 months, IQR 14–59). Prespecified clinical–radiomic and Δ-radiomic signatures were associated with progression-free survival, including B2 = ΔBusyness × Ki-67 (HR 0.38, 95% CI 0.19–0.76, p = 0.006). For progression prediction, the Δ-radiomic model achieved the strongest discrimination, with a nested cross-validation AUC of 0.85 and leave-one-center-out AUC of 0.87. For higher-grade disease, radiomic models also demonstrated high discrimination, with AUCs up to 0.93. Conclusions. Radiomics-derived shape and texture features, especially when combined with clinical markers, may noninvasively identify aggressive PNET phenotypes and support preoperative risk stratification. Prospective validation in larger multicenter cohorts is warranted. Full article

Background/Objectives: Children with Autism Spectrum Disorder (ASD, here on termed autistic children) exhibit motor difficulties in social and non-social contexts. Although previous studies have reported behavioral and neural characteristics, their relationship remains largely unexplored. The current study aimed to investigate the behavioral and neural mechanisms underlying interpersonal synchrony in autistic children using simultaneous kinematic and Functional Near-Infrared Spectroscopy (fNIRS) recordings. Methods: Fifty-eight autistic or non-autistic children participated (mean age = 10.1, standard error = 0.3). fNIRS and an inertial measurement unit were used simultaneously to record the neural activity over frontotemporal and parietal regions and arm movement kinematics during a reach-to-clean-up task across three conditions: Watch—the child observed the tester clean up the blocks; Do—the child cleaned up the blocks independently; and Together—the child and tester cleaned up the blocks synchronously. Results: Behaviorally, autistic children demonstrated longer movement displacement, higher average velocity and acceleration, and a greater number of movement units. In terms of cortical activation, autistic children showed hypoactivation in the bilateral precentral gyrus and right inferior parietal lobe, along with hyperactivation in the right middle frontal gyrus, left inferior frontal gyrus, and left inferior parietal lobule. Correlations between kinematic and neural measures suggest that autistic children rely more on online/feedback control to compensate for reduced feedforward control. Conclusions: This study reveals unique compensatory strategies in autistic children, highlighting the connections between neural and behavioral characteristics. These findings have strong potential to inform the development of ASD screening tools and to guide targeted intervention strategies. Full article

Background: Hypoplastic left heart syndrome (HLHS) is defined as “a spectrum of congenital cardiovascular malformations with normally aligned great arteries without a common atrioventricular junction, characterized by underdevelopment of the left heart with significant hypoplasia of the left ventricle including atresia, stenosis, or hypoplasia of the aortic or mitral valve, or both valves, and hypoplasia of the ascending aorta and aortic arch”. Without treatment, HLHS is usually lethal in the neonate. Many hypotheses have been advanced to explain the etiology of HLHS; however, no single theory appears to fully explain the phenotypic variability seen in HLHS. Furthermore, many of these theories offer no explanations regarding the precipitating events which lead to the development of HLHS. Objective: This review considers and critically evaluates the strengths and weaknesses of the leading theories proposed to explain the pathogenesis of HLHS—including hemodynamic disturbances, primary myocardial structural defects, valvar malformations, and genetic or epigenetic alterations that may provoke developmental and anatomic abnormalities. After presenting each model, we propose a novel, comprehensive, and data-driven framework which may assist researchers in developing models for the pathogenesis of the various subtypes of HLHS. Methods: Key findings from human fetal imaging, histopathology, genetic studies, and animal models were considered, as well as the hypothetical contribution of each in observed HLHS phenotypes. The rationales for these findings as causal factors initiating individual HLHS patterns, as well as how they might contribute to HLHS in general, were critically analyzed. Results: The flow theory is strongly supported by animal models and in utero interventions that demonstrate the impact of altered hemodynamics on cardiac morphogenesis. However, the flow theory fails to identify initial causes of disturbed flow or related histological features of HLHS like endocardial fibroelastosis. The myocardial and valve-first models suggest an important role in developmental defects, but do not necessarily have a strong experimental basis that provides explanations for how they mediate HLHS. Genetic studies in patients with HLHS have identified several candidate causal mutations. However, such genetic causes of HLHS exhibit incomplete phenotypic penetrance and clinical impact. A multifactorial framework attempts to integrate these diverse mechanisms and may provide the most coherent explanation that can accommodate the heterogeneity and variable presentation of HLHS. Such a framework may identify multiple forces that drive disease but does not provide useful pathways for future research about HLHS. Conclusions: No single hypothesis has fully explained how HLHS is initiated, progresses, and presents with the clinical conditions that are encountered by cardiac surgeons and cardiologists. The most current models suggest that the spectrum of HLHS reflects acomplex interaction between genetic susceptibility, flow-dependent cardiac remodeling, and environmental factors in utero. A multifactorial model integrates these diverse mechanisms and may provide the most coherent explanation for the various phenotypic variations in HLHS. Based on our analysis of the most current data and the strengths and weaknesses of the current theoretical frameworks, we propose a novel research strategy aimed at identifying specific cardiac progenitor cell populations whose dysregulation may represent a unifying explanation for the etiology of the various phenotypes of HLHS. Based on the arguments made throughout this manuscript that evaluate the various genetic, structural, and hemodynamic models of initiation of disease, we believe that the significant phenotypic variability across the spectrum of HLHS (i.e., the different anatomic subtypes for “classic” HLHS) most likely reflects different underlying etiologies and mechanisms. At the very least, it is very likely that the timing of the insult is critical in determining anatomic subtype. Based on the published data and the arguments within this manuscript, it seems naive to think that there is a single unifying mechanism explain all forms of HLHLS. Full article

The removal of sulfur compounds such as ethyl mercaptan from natural gas remains a critical challenge due to their detrimental effects on downstream processes, catalyst poisoning, and environmental emissions. In this study, a series of halloysite-supported transition metal oxide catalysts was synthesized and evaluated for the removal of sulfur compounds from natural gas at 25 °C, 200 psi, and 36 mL/min, using 0.5 g of the catalyst. The nanotubular structure and dual surface chemistry of halloysite promote enhanced metal dispersion and improved mass transfer. Single-metal (manganese, copper, zinc, and nickel) catalysts were developed and tested, after which a multi-metal oxide (base) catalyst comprising a composite of the single metals (Zn-Cu-Mn-Ni) was developed as a base catalyst to combine adsorption-active and redox-active functionalities, and its performance was further enhanced by the addition of palladium as promoter. A combination of analytical techniques, including X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), Fourier transform infra-red spectroscopy (FTIR), Brunauer–Emmett–Teller (BET) analysis, scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS), provided evidence that highly dispersed metal oxide phases were formed and the halloysite structure was preserved. XPS data showed the presence of oxidation states of metals that were active (Zn²⁺, Cu²⁺, Ni²⁺, Mn³⁺/Mn⁴⁺ and Pd²⁺), an indication of a redox-active surface for sulfur interaction. Results from the breakthrough experiments showed that the base catalyst significantly improved sulfur removal compared to single-metal catalysts, while the Pd-promoted catalyst exhibited the highest performance, with a breakthrough time of 630 min. Palladium was incorporated at low loading as a promoter, enhancing adsorption performance while maintaining a favorable balance between efficiency and material cost. This enhancement is attributed to synergistic interactions between adsorption-active sites and redox-active species, as well as improved electron transfer facilitated by palladium. The results demonstrate that rational design of multi-metal oxide catalysts supported on naturally occurring halloysite provides an effective and scalable approach for sulfur removal from natural gas, with strong potential for industrial applications. Full article

Neurodegenerative diseases, including Alzheimer’s disease (AD), Parkinson’s disease (PD), and amyotrophic lateral sclerosis (ALS; Lou Gehrig’s disease), represent a growing global health burden characterized by progressive neuronal loss and functional decline. Despite decades of intensive research, effective disease-modifying therapies remain limited, underscoring the urgent need for innovative therapeutic strategies. This review highlights recent advances in the understanding of disease etiology and emerging treatment approaches, with a particular focus on modalities with translational potential. We discussed novel disease-modifying interventions, including gene and cell therapies, RNA-targeting strategies, and immunotherapies aimed at clearing misfolded proteins such as amyloid-β, tau, and α-synuclein. In parallel, we examined the evolving recognition of neuroinflammation and mitochondrial dysfunction as actionable therapeutic targets, alongside progress in precision medicine and biomarker-guided approaches that enable early diagnosis and individualized treatment. Additionally, we summarized developments in repurposed pharmacological agents, neuroprotective compounds, and lifestyle interventions, emphasizing the importance of integrative, multimodal strategies. Across AD, PD, and ALS, convergent molecular mechanisms, including protein misfolding, oxidative stress, and disrupted proteostasis, present opportunities for cross-disease therapeutic targeting. Finally, we addressed key challenges and future directions, including translating preclinical efficacy into clinical success, optimizing CNS-targeted delivery systems, and navigating ethical considerations surrounding gene editing and stem cell therapies. Full article

Implementation remains a central challenge in urban policy, yet the knowledge formats designed to bridge the gap between policy goals and on-the-ground delivery remain under-examined. This study treats 250 UN-Habitat Best Practice reports not as proof of effectiveness but as a standardized genre through which local interventions are narrated, compressed, and made portable for replication. We extract three focal sections, namely Results, Lessons Learned, and Transferability, apply systematic thematic coding with 906 open codes consolidated into axial categories, and compute co-occurrence networks using Jaccard similarity and Lift to detect thematic bundles, holes, and silos within and across sections. Three findings emerge. First, the reporting repertoire narrows progressively, as mean thematic richness declines by 28.2% from Results to Transfers while concentration increases 4.2 times, with substantive dimensions such as governance, equity, sustainability, and evidence losing prevalence to circulation-oriented themes. Second, formal bundle detection yields zero qualifying pairs across all six matrices, indicating a loosely coupled reporting grammar anchored by generic silos rather than integrated implementation packages. Third, structural holes concentrate at the pipeline’s end, where infrastructure transfer and sustainability as transferable value are the most systematically disconnected themes. These patterns reveal a portability paradox in which the reporting format achieves institutional legibility, making practices comparable within a shared vocabulary, but progressively filters out the physical, evidentiary, and context-sensitive content that operational reproduction would require. Full article

Accurate in-season rice (Oryza sativa L.) yield prediction is crucial for improved nitrogen management and climate-smart decision making, yet rigorous comparative benchmarking of machine learning (ML) models using multi-temporal UAV spectral data with independent temporal validation remains limited. This study systematically evaluated four ML algorithms (Random Forest, XGBoost, Neural Network, and Linear Regression) and two Bayesian model averaging ensembles for rice yield prediction using UAV multispectral imagery. Field experiments spanning three growing seasons (2023–2025) at Louisiana State University comprised 9–10 varieties and six nitrogen rates (0–235 kg N ha⁻¹; 576 plots). Vegetation indices and spectral bands from three growth stages were extracted as predictors. Models were compared using 300 random train–test iterations with systematic hyperparameter optimization, followed by independent validation on 2025 data. Among the individual models, XGBoost achieved the highest internal accuracy (R² = 0.87, RMSE = 0.85 t ha⁻¹), substantially outperforming Linear Regression (R² = 0.66, RMSE = 1.32 t ha⁻¹), while reduced BMA reached R² = 0.89. XGBoost demonstrated robust temporal generalization (R² = 0.62, NRMSE = 8.47%) despite environmental variation. The Enhanced Vegetation Index and Normalized Difference Red Edge at 90 days after planting (reproductive stage) were the most stable predictors across 300 iterations. Tree-based ML models substantially outperform traditional linear approaches, providing reliable pre-harvest yield forecasting for operational precision rice production. Full article

Background: Artificial intelligence (AI) is reshaping oncology at every stage of the cancer care pathway, from population-level screening through molecular diagnosis, treatment planning, and post-treatment surveillance. Despite an exponential growth in AI oncology publications exceeding 5000 peer-reviewed studies annually, a critical and persistent gap separates demonstrated algorithmic performance from genuine patient benefit. Most published evidence derives from retrospective, single-institution studies conducted in curated dataset environments that systematically differ from real-world clinical deployment conditions. This comprehensive review examines the translational maturity of AI applications across 18 major malignancies, providing an evidence-stratified, cross-cancer assessment of where AI has fulfilled, approaches, or remains far from fulfilling its transformative potential in oncological care. Methods: A structured narrative review was conducted across PubMed/MEDLINE, Embase, IEEE Xplore, and the Cochrane Library, supplemented by regulatory grey literature including FDA 510(k) decision summaries, CE Technical Files, and ClinicalTrials.gov. Search terms combined cancer site-specific terminology with AI methodology terms and translational outcome descriptors. Studies were only included if they applied an AI or machine learning methodology to a defined clinical oncological task, reported a clearly specified performance evaluation, and involved human subjects or human-derived clinical data. Evidence quality was assessed using QUADAS-2, PROBAST, and Cochrane RoB 2. A five-tier translational readiness framework, grounded in the NIH T0–T4 translational spectrum and CONSORT-AI/SPIRIT-AI guidelines, was applied a priori to enable cross-cancer comparison. A rigorous distinction was maintained between diagnostic accuracy and clinical utility, defined as demonstrated impact on clinical decision-making or patient-centered outcomes. Results: Across all 18 malignancies, AI development varied profoundly by cancer type. Breast cancer and prostate cancer (Tier 1) represent the most mature AI ecosystems, with multiple FDA-cleared tools for mammographic screening and digital pathology achieving prospective multi-institutional validation; however, randomized evidence demonstrating reduced cancer-specific mortality remains absent. Lung, hepatocellular, and melanoma AI (Tier 2) have achieved regulatory milestones but face documented performance disparities across demographic subgroups, including DermaSensor’s 20.7% specificity in primary care settings and HCC model failures in non-viral disease etiologies. Colorectal, glioma, pancreatic, and ovarian cancers (Tier 3) exhibit technical maturity without clinical clarity: colorectal CADe systems increase adenoma detection but meta-analyses of 18,232 patients across 21 RCTs fail to demonstrate improvement in advanced neoplasia detection or cancer incidence reduction. A full study-level presentation of pooled estimates, confidence intervals, and heterogeneity statistics for each cited randomized evidence base across all cancer types would extend beyond the intended scope and format of this cross-cancer narrative review. Gastric, esophageal, cervical, bladder, head and neck, and endometrial cancers (Tier 4) demonstrate promising single-institutional or geographically restricted results without multi-institutional external validation, particularly notable for cervical cancer AI’s transformative potential in low- and middle-income countries constrained by absent regulatory frameworks. Hematologic malignancies, sarcoma, and pediatric solid tumors (Tier 5) face structural barriers, workflow incompatibility in hematopathology, extreme rarity in sarcoma (>70 subtypes, <15,000 US cases annually), and irreducible ethical constraints in pediatric data governance, that cannot be resolved through algorithmic refinement alone. Conclusions: Oncological AI has not yet fulfilled its clinical promise. Across all five translational tiers, a single finding is consistent: diagnostic accuracy is not a surrogate for patient benefit. AI tools with high sensitivity and specificity have repeatedly failed to demonstrate equivalent reductions in cancer-specific mortality, overdiagnosis, or procedural harm under real-world outcome scrutiny. Simultaneously, documented performance disparities across races, ethnicity, disease etiology, and geographic setting reveal that current AI systems risk amplifying the very health inequities they are positioned to resolve. Bridging this translational gap requires three coordinated systemic shifts: regulatory frameworks mandating post-market outcome surveillance as a condition of clinical clearance; prospective trial designs measuring patient-centered endpoints rather than diagnostic concordance alone; and sustained infrastructure investment in federated data governance, demographically inclusive training datasets, and LMIC-accessible regulatory pathways. AI holds genuine potential to reduce cancer mortality on a global scale—but only if held to the evidentiary and equity standards that the stakes of oncological care demand. Full article

Diaphragm dysfunction that leads to respiratory failure is a significant clinical consequence of sepsis-induced critical illness. Diaphragm muscle weakness contributes to morbidity and mortality in these individuals in part due to impaired mitochondrial function. Restoring normal mitochondrial biogenesis is associated with improved survival and physical function. Therefore, identifying reliable biomarkers of mitochondrial dysfunction in diaphragm muscle will allow for more focused and targeted interventions designed to improve the morbidity of critically ill patients. We used a rodent cecal-ligation and puncture (CLP) model to mimic a moderate grade of sepsis. The diaphragm muscle was harvested from adult mice 48 h following CLP (n = 6) or a sham CLP procedure (n = 6). Our primary finding was that moderate grade CLP increases expression of mitochondria-associated microRNA in the diaphragm. Correspondingly, genes associated with mitochondrial biogenesis decreased. Our study provides evidence for sepsis-mediated dysregulation of mitochondrial homeostasis. This may play a role in diaphragm muscle dysfunction, respiratory failure, and difficult weaning from mechanical ventilation in sepsis-induced critical illness. Full article

Background: Negative public attitudes promote undesirable stereotypes and stigma in stutterers. Method: To mitigate negative attitudes, 403 respondents combined from 16 international samples filled out the Public Opinion Survey of Human Attributes–Stuttering (POSHA–S) before and after interventions to improve attitudes and were compared to 249 respondents from seven control groups. Investigators aimed (a) to replicate an extreme case of regression to the mean (i.e., “crossover” effect) reported earlier in larger combined samples in which respondents with high pre-scores ended with low post-scores, respondents with low pre-scores finished with high post-scores, and intermediate scorers were unchanged; and (b) to identify individual POSHA–S items related to overall attitude change and among the high and low scorers. Results: As in previous studies, stuttering attitudes improved in the intervention group but not in the control group. Intervention and control respondents demonstrated “crossover” but less than the earlier samples due to lower pre–post correlations. Item contributions to pre–post change and differences among the three change groups were inconsistent; however, high agreement items by respondents were less likely to vary than low agreement items. Conclusion: The “crossover” effect was replicated, and future research should explore its presence in other measures or conditions. Full article

Background: Dysphagia is a medically complex condition that often necessitates modified food textures to ensure safe swallowing. As smoothies continue to grow in popularity, developing nutritionally balanced recipes that meet the International Dysphagia Diet Standardisation Initiative (IDDSI) Level 3 guidelines may offer practical, appealing options for caregivers and individuals managing dysphagia. Standardized recipes can potentially also support consistency in preparation. Purpose: The purpose of this research was to develop and evaluate palatable smoothie recipes that meet the IDDSI Level 3 consistency guidelines. Method: In this descriptive pilot study, using a pre-test/post-test design, 32 preprofessional students evaluated three smoothies prepared in a laboratory setting. Both fresh and frozen ingredients were used, and each smoothie was tested for IDDSI Level 3 consistency using the IDDSI funnel. Participants rated the smoothies on color, aroma, texture, flavor, appearance, palatability, and overall acceptability using a five-point Likert scale. Results: The results varied across evaluation criteria. Texture and color were the most influential factors in participants’ assessments. The strong impact of texture was an unexpected finding, as all smoothies met Level 3 standards according to the IDDSI funnel. Conversely, the influence of color was expected, as visual presentation is known to significantly affect food perception and acceptance. Conclusions: Given their nutritional value and ease of preparation, smoothies can be a practical addition to modified diets. While IDDSI Level 3 appears to be an appropriate consistency for this purpose, further research may be needed to evaluate the reliability of the IDDSI funnel in ensuring consistent texture outcomes. Full article

Background: Despite remarkable progress in clinical management and screening, colorectal cancer (CRC) remains a major cause of cancer-related deaths worldwide. Sadly, both the number of CRC incidences and the mortality rate are trending upwards, particularly in younger individuals. There is an urgent need for the identification of reliable diagnostic biomarkers and therapeutic targets, and the development of accurate algorithms to guide therapeutic decision-making at the point of care. Here, we leverage multi-model integrative Machine Learning (ML) algorithms using RNA-Seq and somatic mutation data for the classification of tumor–normal samples and the discovery of potential biomarkers and therapeutic targets. Methods: We used RNA sequencing (RNA-Seq) and somatic mutation data from The Cancer Genome Atlas (TCGA) for the development of classification models and the discovery of biomarkers and therapeutic targets. The models were validated using two independent datasets. Results: ML algorithms accurately classified tumor samples and identified a signature for 58 genes, which could serve as potential diagnostic biomarkers. Functional analysis revealed the Wnt and GPCR signaling pathways enriched for somatic mutations. Conclusions: Multi-model integrative ML algorithms integrating gene expression with somatic mutation data represent a powerful approach to the classification of tumor samples and the discovery of biomarkers. Full article