SEARCH · PubMed Health

Results for “machine learning algorithms”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 91 records · Page 5Linked to original sources

PubMed2026

A unified benchmark of supervised and retrieval-based methods for viral genomic sequence classification.

The rapid growth of genomic sequencing demands fast, accurate, and scalable analysis methods. In viral genomic classification, expanding labeled reference collections can make supervised models costly to update and dependent on fixed label sets, motivating retrieval-based genomic classification as a simpler, more flexible alternative. We present a unified benchmark of supervised and retrieval-based methods for viral genomic sequence classification across three viral classification tasks: hepatitis C virus (HCV) genotyping, COVID-19 discrimination, and human papillomavirus (HPV) genotyping. We compare standard sequence encodings (one-hot, k-mers, FCGR) with dense embeddings (dna2vec, DNABERT). For each representation, we evaluate supervised classifiers (Random Forest, Decision Tree, XGBoost) and retrieval-based classification, where sequence vectors are indexed with FAISS and labels are assigned via similarity-weighted k-NN. Furthermore, we benchmark multiple FAISS index types (Flat, IVF, HNSW, IVFPQ, OPQ) to characterize accuracy-speed-memory trade-offs at scale. The results show that XGBoost and retrieval using Flat or IVF indexes achieve strong classification performance under different computational profiles. Compressed indexes such as IVFPQ and OPQ substantially reduce memory usage, although their accuracy loss depends on the dataset and representation. Overall, supervised XGBoost provides a favorable accuracy-size trade-off, while retrieval-based classification remains competitive and allows labeled reference sequences to be incorporated without retraining a global classifier. This benchmark provides practical guidance for selecting sequence representations, classifiers, and vector-search indexes under different accuracy, memory, and update requirements.

Genome, Viral↗

PubMed2025

Federated learning for the pathogenicity annotation of genetic variants in multi-site clinical settings.

MOTIVATION: Rare diseases collectively affect 5% of the population. However, fewer than 50% of rare disease patients receive a molecular diagnosis after whole genome sequencing. Supervised machine learning is a valuable approach for the pathogenicity scoring of human genetic variants. However, existing methods are often trained on curated but limited central repositories, resulting in poor accuracy when tested on external cohorts. Yet, large collections of variants generated at hospitals and research institutions remain inaccessible to machine-learning purposes because of privacy and legal constraints. Federated learning (FL) algorithms have been recently developed enabling institutions to collaboratively train models without sharing their local datasets. RESULTS: Here, we present a proof-of-concept study evaluating the effectiveness of FL for the clinical classification of genetic variants. A comprehensive array of diverse FL strategies was assessed for coding and non-coding Single Nucleotide Variants as well as Copy Number Variants. Our results showed that federated models generally achieved comparable or superior performance to traditional centralized learning. In addition, federated models reached a robust generalization to independent sets with smaller data fractions as compared to their centralized model counterparts. Our findings support the adoption of FL to establish secure multi-institutional collaborations in human variant interpretation. AVAILABILITY AND IMPLEMENTATION: All source code required to reproduce the results presented in this article, implemented in Python, is available under the GNU General Public License v3 at https://github.com/RausellLab/FedLearnVar.

Humans↗

PubMed2026

Inclusion of Multi-Omic Biomarkers Improves Prediction Accuracy of Response, Relapse, and Overall Survival in Acute Myeloid Leukemia Patients Receiving High-Intensity Induction Chemotherapy.

BACKGROUND: Despite advancements in genetic markers for acute myeloid leukemia (AML) risk stratification, outcome prediction remains challenging due to disease heterogeneity and dynamic genetic changes, highlighting the need for reliable biomarkers to improve AML treatment strategies and patient outcomes. To refine outcome predictions, we investigated the use of microbial-derived biomarkers to predict composite complete remission (CRc), relapse, and survival for patients on high- and low-intensity regimens, and to integrate those variables into the widely clinically utilized European Leukemia Network (ELN-2022) genetic risk classification model for high-intensity-treated patients. METHODS: We first developed machine learning models that integrate baseline fecal metabolomics, 16S rRNA-based stool microbiome features, and clinical metadata (sex, antibiotic administration, AML somatic mutations, and cytogenetics) from two cohorts of AML patients (n = 83) undergoing remission induction chemotherapy. Univariate tests and sparse canonical correlation analysis were employed for variable selection and to explore fecal metabolite-microbe relationships. A robust machine learning approach using XGBoost was employed, with 100 stratified data splits (80% training, 20% testing) and coarse-to-fine hyperparameter optimization. Variable importance was aggregated across all models to select key predictors. RESULTS: For high-intensity-treated patients, XGBoost models achieved aggregated AUROC scores of 0.719, 0.729, and 0.65 for CRc, relapse, and overall survival, respectively. For low-intensity-treated patients, these models achieved aggregate AUROC scores of 0.945, 0.724, and 0.768 for these same outcomes, respectively. Integrating the biomarkers identified in the high-intensity machine-learning models with the current ELN-2022 AML risk stratification system effectively stratified patients into risk categories, which obtained higher concordance indices and likelihood ratios, demonstrating improved prognostic accuracy for each outcome compared to ELN-2022 alone. CONCLUSIONS: The inclusion of microbial-derived biomarkers serves as a robust prognostic tool to improve outcome prediction in AML patients, highlighting the potential of its integration into AML risk assessment and paving the way for personalized treatment strategies and improved patient outcomes.

Humans↗

PubMed2026

An Integrated Machine Learning and Genomic Framework for Precise Detection of Gastric Cancer.

This study presents a novel integrative approach for the analysis of high-dimensional gene expression data, leveraging the complementary strengths of unsupervised clustering and supervised classification. Using K-means clustering, the data set is stratified into three distinct clusters, revealing intrinsic biological patterns and relationships. The resulting cluster assignments are subsequently used as pseudolabels to train machine learning models, including support vector machines, random forest, and a stacking ensemble classifier. To validate and enhance the robustness of clustering, complementary methods, such as hierarchical clustering and density-based spatial clustering of applications with noise (DBSCAN), are used, with results visualized through principal component analysis-driven dimensionality reduction. The high predictive accuracy achieved by the classifiers underlines the separability and reliability of the identified clusters. Furthermore, feature importance analysis highlighted key genetic determinants within each cluster, offering actionable insights into potential biomarkers and critical genomic features. This framework bridges the gap between exploratory unsupervised learning and predictive supervised modeling, providing a scalable and interpretable method for analyzing complex genomic data sets. Its applicability extends to biomarker discovery, patient stratification, and other precision medicine applications, emphasizing its utility in advancing genomic research and clinical practice.

Humans↗

PubMed2026

CAKR: commutative algebra k-mer representations for genomics.

Despite the availability of various sequence analysis models, comparative genomic analysis remains a challenge in genomics, genetics, and phylogenetics. Commutative algebra, a fundamental tool in algebraic geometry and number theory, has rarely been used in data and biological sciences. In this study, we introduce commutative algebra k-mer representations as a nonlinear algebraic framework for analyzing genomic sequences. This representation bridges commutative algebra, algebraic topology, combinatorics, and machine learning to establish a mathematical framework for comparative genomic analysis. We evaluate its effectiveness on three tasks including genetic variant classification, phylogenetic tree reconstruction, and viral classification, typically requiring alignment-based, alignment-free, and machine-learning approaches, respectively. In this work, we show that commutative algebra k-mer representations outperform five state-of-the-art sequence analysis methods across twelve primary datasets, with two additional supplementary fragment-placement benchmarks, especially in viral classification, and maintain relatively stable predictive accuracy as dataset size increases, underscoring scalability and robustness.

Genomics↗

PubMed2025

Triage and workflow optimization with artificial intelligence in pediatric imaging.

Artificial intelligence (AI) is being increasingly utilized in various aspects by the radiology department. With an ever-increasing burden on the healthcare system, particularly in emergency units, the need to incorporate AI in patient triage and workflow optimization cannot be overstated. Machine learning (ML)-based algorithms form the core of AI-based software, aiding healthcare professionals at nearly every step in delivering appropriate patient care. Regarding the radiology section of the hospital, AI-based algorithms have proven exceptionally useful in assisting radiologists and technicians with image acquisition. From accurate clinical referrals to scheduling computed tomography/magnetic resonance imaging scan appointments, from ensuring the lowest radiation exposure to offering timely follow-up reminders, ML-based software has indeed revolutionized the concept of modern image acquisition, especially in the pediatric radiology section. Although the implementation of these algorithms is swift, several technical challenges and the limited availability of pediatric datasets preclude their widespread use. The utility of multimodal pediatric datasets, which combine imaging, genomics, and clinical data, for comprehensive AI triage models can help AI systems evolve toward greater adaptability and integration, resulting in enhanced efficiency, reduced turnaround times, and improved patient outcomes in pediatric radiology departments in the future. In this article, we highlight and review the utility of AI and machine learning-based algorithms in efficiently aiding triage and streamlining the workflow in the pediatric radiology section, thereby ensuring an overall improvement in the departmental workflow.

Triage↗

PubMed2026

MetaFX: feature extraction from whole-genome metagenomic sequencing data.

MOTIVATION: Microbial communities consist of thousands of microorganisms and viruses and have a tight connection with an environment, such as gut microbiota modulation of host body metabolism. However, the direct relationship between the presence of certain microorganism and the host state often remains unknown. Toolkits using reference-based approaches are limited to microbes present in databases. Reference-free methods often require enormous resources for metagenomic assembly or results in many poorly interpretable features based on k-mers. RESULTS: Here we present MetaFX-an open-source library for feature extraction from whole-genome metagenomic sequencing data and classification of groups of samples. Using a large volume of metagenomic samples deposited in databases, MetaFX compares samples grouped by metadata criteria (e.g. disease, treatment, etc.) and constructs genomic features distinct for certain types of communities. Features constructed based on statistical k-mer analysis and de Bruijn graphs partition. Those features are used in machine learning models for classification of novel samples. Extracted features can be visualized on de Bruijn graphs and annotated for providing biological insights. We demonstrate the utility of MetaFX by building classification models for 590 human gut samples with inflammatory bowel disease. Our results outperform the previous research disease prediction accuracy up to 17%, and improves classification results compared to taxonomic analysis by 9±10% on average. AVAILABILITY AND IMPLEMENTATION: MetaFX is a feature extraction toolkit applicable for metagenomic datasets analysis and samples classification. The source code, test data, and relevant information for MetaFX are freely accessible at https://github.com/ctlab/metafx under the MIT License. Alternatively, MetaFX can be obtained via http://doi.org/10.5281/zenodo.16949369.

Metagenomics↗

PubMed2026

CCNA2 orchestrates the PI3K/AKT signaling axis to propel prostate cancer metastasis.

BACKGROUND: Prostate cancer (PCa) remains one of the most common malignancies in men, posing a persistent global burden in terms of both public health and socioeconomic costs. Although early detection is essential for improving patient outcomes, existing clinical tools, including prostate-specific antigen (PSA) screening, digital rectal examination, and transrectal ultrasound-guided biopsy, are hampered by suboptimal specificity and positive predictive value, resulting in frequent overdiagnosis and overtreatment of indolent lesions while missing a subset of aggressive tumors at an early stage. In this context, the rapid advancement of high-throughput omics technologies, coupled with sophisticated machine learning (ML) algorithms, provides a powerful computational framework to dissect high-dimensional genomic data, uncover latent gene expression signatures, and identify candidate biomarkers with superior discriminative performance over conventional clinicopathological parameters. Therefore, in this study, we sought to screen for crucial ML-based biomarkers associated with PCa, with a particular focus on systematically assessing the diagnostic and prognostic value of CCNA2. Leveraging large-scale transcriptomic cohorts from public repositories, we employed an ensemble of ML approaches to prioritize candidate genes and subsequently evaluated the diagnostic performance of CCNA2 through receiver operating characteristic curve analysis, as well as its prognostic utility via Kaplan-Meier survival estimation and multivariate Cox proportional hazards modeling. Our findings are anticipated to elucidate the molecular landscape of PCa and offer a promising biomarker candidate for early detection and risk stratification. METHODS: This study integrated single-cell RNA sequencing, bulk transcriptomic data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) repositories, immunofluorescence, and multiple ML algorithms with in vitro functional assays to evaluate CCNA2 expression, clinical relevance, and biological behavior in PCa. RESULTS: CCNA2 was linked to metastasis and poor prognosis. High CCNA2 expression significantly correlated with adverse survival outcomes, and knockdown of CCNA2 suppressed proliferation, migration, and invasion in PCa cell lines. Mechanistically, CCNA2 modulated the PI3K/AKT signaling pathway. An ML-based diagnostic model incorporating CCNA2 demonstrated high predictive accuracy across multiple validation cohorts. CONCLUSIONS: CCNA2 serves as a promising prognostic biomarker and therapeutic target in prostate adenocarcinoma, driving tumor progression potentially via the PI3K/AKT axis.

CCNA2↗

PubMed2024

Digital pathology and spatial omics in steatohepatitis: Clinical applications and discovery potentials.

Steatohepatitis with diverse etiologies is the most common histological manifestation in patients with liver disease. However, there are currently no specific histopathological features pathognomonic for metabolic dysfunction-associated steatotic liver disease, alcohol-associated liver disease, or metabolic dysfunction-associated steatotic liver disease with increased alcohol intake. Digitizing traditional pathology slides has created an emerging field of digital pathology, allowing for easier access, storage, sharing, and analysis of whole-slide images. Artificial intelligence (AI) algorithms have been developed for whole-slide images to enhance the accuracy and speed of the histological interpretation of steatohepatitis and are currently employed in biomarker development. Spatial biology is a novel field that enables investigators to map gene and protein expression within a specific region of interest on liver histological sections, examine disease heterogeneity within tissues, and understand the relationship between molecular changes and distinct tissue morphology. Here, we review the utility of digital pathology (using linear and nonlinear microscopy) augmented with AI analysis to improve the accuracy of histological interpretation. We will also discuss the spatial omics landscape with special emphasis on the strengths and limitations of established spatial transcriptomics and proteomics technologies and their application in steatohepatitis. We then highlight the power of multimodal integration of digital pathology augmented by machine learning (ML)algorithms with spatial biology. The review concludes with a discussion of the current gaps in knowledge, the limitations and premises of these tools and technologies, and the areas of future research.

Humans↗

PubMed2026

EPIC: Event Prototyping via Information Constrained graph learning for personalized cancer driver gene prediction.

MOTIVATION: Precision oncology relies on accurately distinguishing patient-specific driver mutations from the vast background of passenger alterations. While graph-based computational methods have emerged as powerful tools for this task, they often struggle to preserve the distinct genomic context of individual mutations within complex biological networks. Consequently, subtle patient-specific driver signals are frequently obscured by dominant topological patterns, critically impeding the identification of individualized oncogenic events essential for personalized cancer therapy. RESULTS: To address this, we propose EPIC, a novel framework for Event Prototyping via Information Constrained Graph Learning. Unlike traditional node-centric approaches, EPIC redefines driver prediction as a metric learning task in an event embedding space. We introduce an information-constrained learning strategy that imposes explicit geometric constraints on feature variance, effectively preventing feature collapse and ensuring that low-frequency driver signals are distinctively preserved. Experiments on large-scale cancer cohorts demonstrate that EPIC significantly outperforms established baselines. Notably, the model prioritizes low-frequency driver variants typically overlooked by population-based methods, mapping them to critical oncogenic mechanisms associated with drug resistance and metastasis. Furthermore, clinical actionability analysis confirms that EPIC substantially expands the patient population eligible for targeted therapies. EPIC provides a robust and context-aware solution for personalized cancer driver discovery, bridging the gap between genomic data and actionable therapeutic insights. AVAILABILITY AND IMPLEMENTATION: The source code and datasets are available at https://github.com/spcho-dev/EPIC.

Humans↗

PubMed2026

Identification of Immune Response-Related Proteomic Biomarkers in Moyamoya Disease Using Serum Olink Proteomics.

Moyamoya disease, a rare chronic cerebrovascular disorder, requires invasive digital subtraction angiography (DSA) for diagnosis. This study employed high-throughput proteomics to identify plasma biomarkers for Moyamoya disease diagnosis. We conducted immunopanel analysis using the Olink platform to evaluate 92 immune-related proteins in plasma samples from 88 Moyamoya disease patients and 88 healthy controls. Key proteins were identified through differential expression analysis, GO, and KEGG enrichment analysis. A diagnostic model was constructed using LASSO regression, Boruta algorithm, and machine learning models including random forest and XGBoost. Validation of these proteins was performed using GEO external data sets, followed by prediction of potential therapeutic drugs and molecular docking validation through pharmacogenomic databases. A total of 44 differentially expressed proteins were identified through the Olink immunopanel, with 12 downregulated and 32 upregulated. GO and KEGG analyses revealed significant enrichment of these proteins in innate immune responses and signaling pathways such as NF-kB and MAPK. Through LASSO, random forest, and protein under-area analysis, four potential biomarkers for Moyamoya disease (MGMT, SIT1, PRDX1, TRAF2) were identified. A diagnostic model using these proteins showed the highest AUC value with the XGBoost model. Additionally, TRAF2 and PRDX1 exhibited significant expression differences in Moyamoya disease patients within the GEO data set. Our study revealed the immune landscape of Moyamoya disease, identified four biomarkers, and established a variety of diagnostic models.

Humans↗

PubMed2025

Predicting natural variation in the yeast phenotypic landscape with machine learning.

Most organismal traits result from the complex interplay of many genetic and environmental factors, making their prediction difficult. Here, we used machine learning (ML) models to explore phenotype predictions for 223 traits measured across 1011 genome-sequenced Saccharomyces cerevisiae strains isolated worldwide. We benchmarked a ML pipeline with multiple linear and non-linear models to predict phenotypes from genotypes and gene expression, and determined gradient boosting machines as the best-performing model. Gene function disruption scores and gene presence/absence emerged as best predictors, suggesting a considerable contribution of the accessory genome in controlling phenotypes. The prediction accuracy broadly varied among phenotypes, with stress resistance being easier to predict compared to growth across nutrients. ML identified relevant genomic features linked to phenotypes, including high-impact variants with established relationships to phenotypes, despite these being rare in the population. Near-perfect accuracies were achieved when other phenomics data mostly in similar conditions were used, suggesting that useful information can be conveyed across phenotypes. Overall, our study underscores the power of ML to interpret the functional outcome of genetic variants.

Genetic Variation↗

PubMed2026

Transcriptome Analysis and Experimental Validation of Palmitoylation- Related Biomarkers in Atherosclerosis.

INTRODUCTION: Protein palmitoylation contributes to membrane localisation, signal transduction, and cell-fate regulation. It is closely associated with lipid metabolic dysfunction, immune inflammation, and vascular remodelling in atherosclerosis (AS). However, key palmitoylation-related transcriptomic markers and their potential causal associations with AS remain incompletely defined. METHODS: The Gene Expression Omnibus (GEO) dataset GSE100927 was used as the training cohort, and GSE43292 was used as an external validation cohort. Differentially expressed genes were identified using limma and intersected with palmitoylation-related genes to obtain palmitoylation-related differentially expressed genes (PRDEGs). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were then performed using clusterProfiler. Two-sample Mendelian randomisation was used to evaluate potential causal relationships between characteristic genes and AS. Feature selection was conducted using random forest and support vector machine recursive feature elimination (SVM-RFE), and the overlapping genes selected by both methods were retained. Receiver operating characteristic (ROC) curves were used to assess diagnostic performance. A five-gene nomogram was constructed, and its clinical utility was evaluated using calibration curves and decision curve analysis (DCA). Gene set variation analysis (GSVA) was applied to compare pathway activity between high- and low-expression groups for each core gene. Single-cell analysis using Seurat and expression-based cell-cell communication analysis using CellChat were conducted with GSE159677, and upstream transcription factors were predicted using NetworkAnalyst. For in vivo validation, an AS model was established in ApoE⁸/⁸ mice fed a high-fat diet, and aortic gene and protein expression were assessed by RT-qPCR and western blotting. RESULTS: In GSE100927, 51 PRDEGs were identified. GO and KEGG enrichment analyses highlighted pathways associated with regulation of monoatomic ion transport, sarcomere and myofibril organisation, and immune inflammation. Mendelian randomisation suggested a potential protective causal association between SLC7A7 and AS. By integrating MR with random forest and SVM-RFE feature selection, we prioritised five core genes: PLCB2, GMIP, NEXN, PLN, and SLC7A7. These genes showed good diagnostic performance in GSE43292. The resulting nomogram was well calibrated and demonstrated stable net benefit in decision curve and clinical impact curve analyses. Single-gene GSVA identified consistently activated pathways across multiple genes, including innate and adaptive immune recognition, calcium signalling and myocardial contraction/cardiomyopathy, extracellular matrix-receptor interaction, cell junction pathways, autophagy-lysosome pathways, and several metabolic programmes. At the single-cell level, PLCB2 and GMIP were predominantly expressed in T cells and macrophages, NEXN and PLN were enriched in vascular smooth muscle cells, and SLC7A7 was mainly expressed in macrophages. CellChat analysis indicated increased signals for immune-related ligand-receptor interactions. In ApoE⁸/⁸ mice fed a high-fat diet, PLCB2, GMIP, and SLC7A7 were upregulated, whereas NEXN and PLN were downregulated; protein-level changes were concordant with the transcriptomic trends. DISCUSSION: These findings indicate that palmitoylation-related dysregulation in AS converges on immune inflammation, calcium signalling/contractile programmes, ECM remodelling, and autophagy-linked metabolism. The five-gene panel is supported by external validation, single-cell localisation to immune and vascular compartments, and concordant results in ApoE⁸/⁸ mice. CONCLUSION: This study identified and validated five palmitoylation-related genes associated with AS. SLC7A7 showed a potential protective causal signal in MR analysis. The enriched pathway patterns linked these genes to immune inflammation, calcium signalling-contraction coupling, ECM remodelling, cell adhesion, and autophagy- associated metabolic reprogramming. The five-gene nomogram showed potential utility for diagnostic classification and decision support, nominating candidate biomarkers and pathway targets for AS molecular subtyping, diagnosis, and mechanistic investigation.

Atherosclerosis (AS)↗

PubMed2026

Machine Learning and Metabolomics to Characterize Warburg-Like Metabolic Subtypes in Human Retinal Endothelial Cells Exposed to Risk Factors Associated With Proliferative Diabetic Retinopathy.

PURPOSE: High glucose (HG), hypoxia (Hyp), and their combination are major risk factors for proliferative diabetic retinopathy (PDR). Although these conditions induce features of the Warburg-like metabolic reprogramming in human retinal endothelial cells (HRECs), it remains unclear whether they produce distinct metabolic and angiogenic subtypes. This study aimed to characterize the Warburg-like-associated metabolic heterogeneity induced by these PDR-related risk factors and evaluate the ability of supervised machine-learning models to distinguish these subtypes. METHODS: HRECs were cultured under normoglycemic, HG, Hyp (2% O2), and combined HG-Hyp conditions. Untargeted LC-MS/MS metabolomics quantified metabolites spanning carbohydrates, amino acids, nucleotides, and lipids. Principal component analysis (PCA) assessed overall metabolic variation, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis identified metabolic pathways associated with angiogenesis. In vitro angiogenesis assays measured endothelial tube formation and branching. Nine supervised classifiers (decision tree, logistic regression, naïve Bayes, random forest, K-Nearest Neighbors, neural network, gradient boosting, AdaBoost, and Support Vector Machine) were trained on the highest-ranked metabolites selected by the Information Gain Ratio feature-ranking approach. Model performance was evaluated using 10-fold cross-validation, leave-one-out cross-validation (LOOCV), permutation testing, and a classifier stability analysis under biologically meaningful distributional shift using an independent chemically induced hypoxia model (CoCl2). RESULTS: PCA revealed partial separation of metabolic profiles across conditions, indicating different Warburg-like metabolic subtypes. The combined HG-Hyp condition exhibited enhanced angiogenic potential relative to either HG or Hyp alone. KEGG pathway enrichment analysis identified fatty acid biosynthesis and elongation among the most significantly enriched pathways in HRECs under combined HG-Hyp conditions, alongside amino sugar and nucleotide sugar metabolism, glycerophospholipid metabolism, the pentose phosphate pathway, and glycolysis/gluconeogenesis. Supervised machine-learning classifiers distinguished these metabolic subtypes, with AdaBoost and gradient Boosting showing the most balanced, reproducible performance across 10-fold cross-validation, LOOCV, and permutation testing, and remaining the most reliable classifiers under domain-shift testing (area under the curve = 0.88, P = 0.0061). CONCLUSIONS: In this exploratory analysis, HG, Hyp, and their combination drive metabolically and functionally distinct subtypes of Warburg-like metabolic reprogramming in HRECs, with HG-Hyp in combination producing a highly angiogenic phenotype. Boosting-based ensemble classifiers provide a promising framework for detecting these subtypes even under domain-shift conditions, warranting validation in larger independent datasets. TRANSLATIONAL RELEVANCE: Integrating metabolomics with machine-learning classification offers a strategy to identify Warburg-like metabolic subtypes in retinal endothelial cells, providing insights into angiogenic mechanisms and guiding the development of targeted diagnostics or therapeutics for PDR.

Humans↗

PubMed2026

Benchmark of biomarker identification and prognostic modeling methods on diverse censored data.

The practices of identifying biomarkers and developing prognostic models using genomic data has become increasingly prevalent. Such data often features characteristics that make these practices difficult, namely high dimensionality, correlations between predictors, and sparsity. Many modern methods have been developed to address these problematic characteristics while performing feature selection and prognostic modeling, but a large-scale comparison of their performances in these tasks on diverse right-censored time to event data (aka survival time data) is much needed. We have compiled many existing methods, including some machine learning methods, several which have performed well in previous benchmarks, primarily for comparison in regards to variable selection capability, and secondarily for survival time prediction on many synthetic datasets with varying levels of sparsity, correlation between predictors, and signal strength of informative predictors. For illustration, we have also performed multiple analyses on a publicly available and widely used cancer cohort from The Cancer Genome Atlas using these methods. We evaluated the methods through extensive simulation studies in terms of the false discovery rate, F1-score, concordance index, Brier score, root mean square error, and computation time. Of the methods compared, CoxBoost and the Adaptive LASSO performed well in all metrics, and the LASSO and elastic net excelled when evaluating concordance index and F1-score. The Benjamini-Hoschberg and q-value procedures showed volatile performances in controlling the false discovery rate. Some methods' performances were greatly affected by differences in the data characteristics. With our extensive numerical study, we have identified the best performing methods for a plethora of data characteristics using informative metrics. This will help cancer researchers in choosing the best approach for their needs when working with genomic data.

Humans↗

PubMed2026

Meta-PseU: A meta-classifier for robust prediction of RNA pseudouridine modification sites from long sequences.

BACKGROUND AND OBJECTIVES: Pseudouridine (Ψ) represents one of the most abundant and conserved RNA modifications. Ψ provides an additional hydrogen-bond donor that enhances RNA structural stability and modulates translation. It participates in diverse biological processes, including RNA-protein interactions, splicing, translational control, and stress responses. Aberrant pseudouridylation is implicated in cancer, neurodegenerative disorders, and autoimmune diseases. Despite its biological importance, experimental identification of Ψ sites remains time-consuming and costly, limiting the feasibility of transcriptome-wide profiling. Computational approaches have therefore become essential complements to experimental techniques. However, state-of-the-art machine-learning and deep-learning predictors often suffer from limited generalizability due to small training datasets. To overcome these issues, we aim at constructing new long-sequence datasets and developing a novel Ψ site predictor. METHODS: New long-sequence datasets were constructed as benchmarks for RNA Ψ-site prediction. The Ψ modification sites in RMBase 3.0 were mapped to the reference genomes across three species of human, mouse, and yeast, and the RNA sequences with a length of 201 were generated by extending the upstream and downstream from the mapped, central sites. To eliminate sequence redundancy, the sequences were clustered using CD-HIT with a 70% sequence identity threshold. We developed Meta-PseU, a logistic regression-based meta-classifier that considered 118 machine learning and deep learning classifiers. The datasets and programs are freely accessible at https://github.com/kuratahiroyuki/MetaPseU. RESULTS: By optimizing model configuration, we proposed the Meta-PseU model stacking 32 machine learning and deep learning classifiers out of 118 classifiers. Meta-PseU substantially improved model generalizability, overcoming a key limitation of existing approaches. It greatly outperformed state-of-the-art predictors and achieved increasing accuracy with increasing sequence length. CONCLUSIONS: Long-sequence datasets were newly constructed as benchmarks for RNA Ψ-site prediction. Meta-PseU offers a new framework for robust Ψ-site identification by using long sequences.

Pseudouridine↗

PubMed2025

Generative AI Models in Time-Varying Biomedical Data: Scoping Review.

BACKGROUND: Trajectory modeling is a long-standing challenge in the application of computational methods to health care. In the age of big data, traditional statistical and machine learning methods do not achieve satisfactory results as they often fail to capture the complex underlying distributions of multimodal health data and long-term dependencies throughout medical histories. Recent advances in generative artificial intelligence (AI) have provided powerful tools to represent complex distributions and patterns with minimal underlying assumptions, with major impact in fields such as finance and environmental sciences, prompting researchers to apply these methods for disease modeling in health care. OBJECTIVE: While AI methods have proven powerful, their application in clinical practice remains limited due to their highly complex nature. The proliferation of AI algorithms also poses a significant challenge for nondevelopers to track and incorporate these advances into clinical research and application. In this paper, we introduce basic concepts in generative AI and discuss current algorithms and how they can be applied to health care for practitioners with little background in computer science. METHODS: We surveyed peer-reviewed papers on generative AI models with specific applications to time-series health data. Our search included single- and multimodal generative AI models that operated over structured and unstructured data, physiological waveforms, medical imaging, and multi-omics data. We introduce current generative AI methods, review their applications, and discuss their limitations and future directions in each data modality. RESULTS: We followed the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines and reviewed 155 articles on generative AI applications to time-series health care data across modalities. Furthermore, we offer a systematic framework for clinicians to easily identify suitable AI methods for their data and task at hand. CONCLUSIONS: We reviewed and critiqued existing applications of generative AI to time-series health data with the aim of bridging the gap between computational methods and clinical application. We also identified the shortcomings of existing approaches and highlighted recent advances in generative AI that represent promising directions for health care modeling.

Artificial Intelligence↗

PubMed2026

A machine learning-derived intratumoral heterogeneity-related signature predicts the prognosis for and therapeutic response in patients with skin cutaneous melanoma.

BACKGROUND: Reliable biomarkers for predicting prognosis and therapeutic response in skin cutaneous melanoma (SKCM) remain limited. This study aimed to develop an intratumoral heterogeneity (ITH)-related prognostic signature for SKCM using integrative machine learning. METHODS: RNA sequencing (RNA-seq) data from 472 SKCM patients in The Cancer Genome Atlas (TCGA) and 214 patients in the GSE65904 cohort were analyzed. ITH scores were calculated using the DEPTH2 algorithm. Differentially expressed genes (DEGs) were identified between high- and low-ITH groups [|log2fold change (FC)| ≥1, false discovery rate (FDR) <0.05]. Based on 38 prognostic DEGs identified by univariate Cox regression, we employed an integrative framework of 101 machine learning algorithm combinations to construct prognostic models in the TCGA training cohort. The model with the highest average concordance index (C-index) was validated in the GSE65904 cohort and selected as the prognostic ITH-related signature (PIRS). Associations of the PIRS risk score with tumor mutational burden (TMB), immune cell infiltration, immune checkpoint gene expression, and drug sensitivity were systematically evaluated. Model performance was assessed using receiver operating characteristic (ROC) curves and Cox regression analyses. RESULTS: A 38-gene PIRS was constructed using the plsRcox algorithm. Patients with high PIRS risk scores exhibited significantly poorer overall survival (OS) in both the TCGA and Gene Expression Omnibus (GEO) cohorts. The PIRS was identified as an independent prognostic factor, with area under the curve (AUC) values of 0.779, 0.734, and 0.756 for 1-, 3-, and 5-year survival, respectively. High-risk samples displayed significantly lower TMB (P<0.05), reduced immune and stromal cell infiltration (P<0.001), downregulated immune function, and decreased expression of immune checkpoint genes. Additionally, high- and low-PIRS risk score groups exhibited distinct sensitivity patterns to different classes of targeted agents. CONCLUSIONS: The machine learning-derived PIRS robustly predicts prognosis in SKCM patients. Its clinical application is promising for optimizing patient risk stratification and treatment decisions, though further prospective validation is warranted.

Skin cutaneous melanoma (SKCM)↗