SEARCH · PubMed Health

Results for “Deep Learning”

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 163 records · Page 9Linked to original sources

PubMed2025

DiCARN-DNase: enhancing cell-to-cell Hi-C resolution using dilated cascading ResNet with self-attention and DNase-seq chromatin accessibility data.

MOTIVATION: The spatial organization of chromatin is fundamental to gene regulation and essential for proper cellular function. The Hi-C technique remains the leading method for unraveling 3D genome structures, but the limited availability of high-resolution (HR) Hi-C data poses significant challenges for comprehensive analysis. Deep learning models have been developed to predict HR Hi-C data from low-resolution counterparts. Early Convolutional Neural Network (CNN)-based models improved resolution but struggled with issues like blurring and capturing fine details. In contrast, Generative Adversarial Network (GAN)-based methods encountered difficulties in maintaining diversity and generalization. Additionally, most existing algorithms perform poorly in cross-cell line generalization, where a model trained on one cell type is used to enhance HR data in another cell type. RESULTS: In this work, we propose Dilated Cascading Residual Network (DiCARN) to overcome these challenges and improve Hi-C data resolution. DiCARN leverages dilated convolutions and cascading residuals to capture a broader context while preserving fine-grained genomic interactions. Additionally, we incorporate DNase-seq data into our model, providing a robust framework that demonstrates superior generalizability across cell lines in HR Hi-C data reconstruction. AVAILABILITY AND IMPLEMENTATION: DiCARN is publicly available at https://github.com/OluwadareLab/DiCARN.

Chromatin↗

PubMed2026

DeepPlaque: a scalable multimodal platform for Aβ pathology and cell analysis in Alzheimer's disease.

Histological analysis is essential for understanding disease pathology and the microenvironment, particularly in Alzheimer's disease (AD), characterized by beta-amyloid (Aβ) plaques that exist as diffuse, fibrillar, and core species, with distinct toxicity levels. However, accurate classification of Aβ plaque types in postmortem brain tissues and profiling of surrounding cells present significant challenges. To address these challenges, we developed "DeepPlaque", an integrated system featuring "PlaqueNet", a deep learning model for automated classification of Aβ plaque species from diverse imaging platforms. DeepPlaque includes automated workflows for cellular phenotyping and proteomic profiling through targeted laser microdissection. PlaqueNet achieves expert-level accuracy (AUC > 90%) in classifying the 3 major Aβ plaque species, supporting consistent and large-scale annotation. By integrating spatial cellular phenotyping with laser microdissection, DeepPlaque enables high-throughput proteomic analysis of Aβ plaque niches, revealing that microglia are more abundant around core and fibrillar Aβ plaques, with increased expression of apolipoprotein E and amyloid precursor protein in core Aβ plaques. This customizable platform enhances the molecular and cellular characterization of Aβ plaque-associated environments, providing critical insights into AD pathology.

Alzheimer Disease↗

PubMed2026

Artificial Intelligence in Predicting Systemic Complications From Retinal Findings: A New Frontier in Precision Medicine.

Innovations in retinal imaging technologies and growing evidence from retinal imaging of systemic and neurodegenerative diseases have begun to explore the utility of retinal imaging in diagnosing these conditions. Since the retina shares embryological origins with the central nervous system and reflects systemic microvascular characteristics, it is well positioned for noninvasive observation of patients' systemic and neural health. Moreover, accessibility of retinal imaging has improved with the increasing number of ophthalmology clinics. Rapid improvements in various deep learning (DL) tools have also catalyzed the automation of retinal imaging analysis. Systems that utilize DL for retinal imaging are being developed to assist with disease recognition, clinical judgment, and prognostic assessment of systemic health. Various imaging modalities are being integrated with existing genomic and clinical data to estimate an individual's predisposition to certain conditions. Contrary to many existing reviews, the objective of this review is to synthesize the most recent clinical and technological evidence on DL-based diagnostic systems for retinal imaging, with a focus on how different network architectures and their combinations have been developed, validated, and applied across systemic disease detection and prediction. Specifically, this review examines the datasets, model validation approaches, and automated diagnostic systems reported in recent literature. It discusses the extent to which these advancements address existing barriers toward real-time diagnostic application across clinical disciplines. Integrating retinal imaging with DL is an innovative and promising approach to precision medicine and health risk reduction.

artificial intelligence↗

PubMed2025

A neural network model enables worm tracking in challenging conditions and increases signal-to-noise ratio in phenotypic screens.

High-resolution posture tracking of C. elegans has applications in genetics, neuroscience, and drug screening. While classic methods can reliably track isolated worms on uniform backgrounds, they fail when worms overlap, coil, or move in complex environments. Model-based tracking and deep learning approaches have addressed these issues to an extent, but there is still significant room for improvement in tracking crawling worms. Here we train a version of the DeepTangle algorithm developed for swimming worms using a combination of data derived from Tierpsy tracker and hand-annotated data for more difficult cases. DeepTangleCrawl (DTC) outperforms existing methods, reducing failure rates and producing more continuous, gap-free worm trajectories that are less likely to be interrupted by collisions between worms or self-intersecting postures (coils). We show that DTC enables the analysis of previously inaccessible behaviours and increases the signal-to-noise ratio in phenotypic screens, even for data that was specifically collected to be compatible with legacy trackers including low worm density and thin bacterial lawns. DTC broadens the applicability of high-throughput worm imaging to more complex behaviours that involve worm-worm interactions and more naturalistic environments including thicker bacterial lawns.

Caenorhabditis elegans↗

PubMed2025

Lit-OTAR framework for extracting biological evidences from literature.

SUMMARY: The lit-OTAR framework, developed through a collaboration between Europe PMC and Open Targets, leverages deep learning to revolutionize drug discovery by extracting evidence from scientific literature for drug target identification and validation. This novel framework combines named entity recognition for identifying gene/protein (target), disease, organism, and chemical/drug within scientific texts, and entity normalization to map these entities to databases like Ensembl, Experimental Factor Ontology, and ChEMBL. Continuously operational, it has processed over 39 million abstracts and 4.5 million full-text articles and preprints to date, identifying more than 48.5 million unique associations that significantly help accelerate the drug discovery process and scientific research >29.9 m distinct target-disease, 11.8 m distinct target-drug, and 8.3 m distinct disease-drug relationships. AVAILABILITY AND IMPLEMENTATION: The results are accessible through Europe PMC's SciLite web app (https://europepmc.org/) and its annotations API (https://europepmc.org/annotationsapi), as well as via the Open Targets Platform (https://platform.opentargets.org/). The daily pipeline is available at https://github.com/ML4LitS/otar-maintenance, and the Open Targets ETL processes are available at https://github.com/opentargets.

Drug Discovery↗

PubMed2026

PAT: An Image Analysis Tool for Automated Scoring of Pollen in Alexander-Stained Anthers.

Quantitative pollen viability analysis is a critical but labor-intensive step in plant reproductive biology. Existing deep-learning Segment Anything Models (SAM) fail to reliably segment viable pollen in Alexander-stained anthers. To address this, we fine-tuned an existing Cellpose-SAM model for pollen segmentation. We integrated it into PAT (Pollen Analysis Tool), a cross-platform desktop application. PAT features instance segmentation with interactive quality control, an in-app model retraining module, and publication-ready statistical outputs. We deployed PAT in an EMS suppressor screen of semi-sterile Arabidopsis smg7-6 mutants, enabling efficient candidate prioritization for whole-genome sequencing and mapping of the candidate mutation. This screen led to the identification of a point mutation in CAP-D2 (capd2-2), a Condensin I subunit, that rescues the smg7-6 meiotic phenotype. Notably, mutation in a Condensin II subunits (CAP-D3 and CAP-H2) does not confer rescue. Further characterization suggests the capd2-2 allele is hypomorphic, showing no defects in vegetative growth, chromocenter compaction, or transposable element silencing. Collectively, we demonstrate that accessible AI tools have the potential to bridge gaps in plant phenotyping and accelerate the pace of biological discovery.

Alexander staining↗

PubMed2026

H&E to recurrence score: A step forward, but not yet a substitute for genomic testing.

Shamai and colleagues developed a multimodal deep-learning model that predicts Oncotype DX recurrence scores from routine H&E slides and clinicopathological variables in hormone receptor‑positive, HER2‑negative early breast cancer. Validated across the TAILORx trial and six external cohorts (over 5000 patients), the model achieved an AUC of 0.898 for identifying recurrence score ≥26 and recapitulated genomic assay patterns of chemotherapy benefit. Notably, 31% of clinically high-risk postmenopausal women were downgraded to low risk by AI, suggesting potential to reduce overtreatment. However, several limitations preclude immediate clinical substitution for genomic testing. First, intratumoural heterogeneity leads to discordant predictions with unclear management guidance. Second, the model's chemotherapy benefit estimates rely on TAILORx's age-based menopausal surrogates, which may not reflect real-world hormonal status or LHRH agonist use. Third, predictive value in node-positive disease remains untested in randomised datasets such as RxPONDER. Additionally, calibration uncertainty near risk thresholds and global scalability issues (including IHC requirements and digital pathology infrastructure) persist. While this represents a landmark step toward democratising precision oncology, the AI tool should currently serve as a complementary decision aid, with genomic testing remaining the gold standard for intermediate, borderline, or discordant cases.

Breast cancer↗

PubMed2025

Histology-Based Virtual RNA Inference Identifies Pathways Associated With Metastasis Risk in Colorectal Cancer.

Colorectal cancer (CRC) remains a major health concern, with >150,000 new diagnoses and >50,000 deaths annually in the United States, underscoring an urgent need for improved screening, prognostication, disease management, and therapeutic approaches. The tumor microenvironment (TME)-comprising cancerous and immune cells interacting within the tumor's spatial architecture-plays a critical role in disease progression and treatment outcomes, reinforcing its importance as a prognostic marker for metastasis and recurrence risk. However, traditional methods for TME characterization, such as bulk transcriptomics and multiplex protein assays, lack sufficient spatial resolution. Although spatial transcriptomics (ST) allows for the high-resolution mapping of whole transcriptomes at near-cellular resolution, current ST technologies (eg, Visium and Xenium) are limited by high costs, low throughput, and issues with reproducibility, preventing their widespread application in large-scale molecular epidemiology studies. In this study, we refined and implemented virtual RNA inference (VRI) to derive ST-level molecular information directly from hematoxylin and eosin (H&E)-stained tissue images. Our VRI models were trained on the largest matched CRC ST data set to date, comprising 45 patients and >300,000 Visium spots from primary tumors. Using state-of-the-art deep learning models (UNI, ResNet-50, Vision Transformer, and Vision Mamba), we achieved a median Spearman's correlation coefficient of 0.546 between predicted and measured spot-level expression. As validation, VRI-derived gene signatures linked to specific tissue regions (tumor, interface, submucosa, stroma, serosa, muscularis, and inflammation) showed strong concordance with signatures generated via direct ST, and VRI performed accurately in estimating cell-type proportions spatially from H&E slides. In an expanded CRC cohort controlling for tumor invasiveness and clinical factors, we further identified VRI-derived gene signatures significantly associated with key prognostic outcomes, including metastasis status. Although certain tumor-related pathways are not fully captured by histology alone, our findings highlight the ability of VRI to infer a wide range of "histology-associated" biological pathways at near-cellular resolution without requiring ST profiling. Future efforts will extend this framework to expand TME phenotyping from standard H&E tissue images, with the potential to accelerate translational CRC research at scale.

Humans↗

PubMed2025

Accurate somatic small variant discovery for multiple sequencing technologies with DeepSomatic.

Somatic variant detection is an integral part of cancer genomics analysis. While most methods have focused on short-read sequencing, long-read technologies offer potential advantages in repeat mapping and variant phasing. We present DeepSomatic, a deep-learning method for detecting somatic small nucleotide variations and insertions and deletions from both short-read and long-read data. The method has modes for whole-genome and whole-exome sequencing and can run on tumor-normal, tumor-only and formalin-fixed paraffin-embedded samples. To train DeepSomatic and help address the dearth of publicly available training and benchmarking data for somatic variant detection, we generated and make openly available the Cancer Standards Long-read Evaluation (CASTLE) dataset of six matched tumor-normal cell line pairs whole-genome sequenced with Illumina, PacBio HiFi and Oxford Nanopore Technologies, along with benchmark variant sets. Across samples, both cell line and patient-derived, and across short-read and long-read sequencing technologies, DeepSomatic consistently outperforms existing callers.

Humans↗

PubMed2026

GBFN: A gated bimodal fusion network leveraging foundation model embeddings for cancer drug sensitivity prediction.

Despite recent progress in deep learning for cancer drug sensitivity prediction, many existing models still rely on task-specific representation learning or relatively simple multimodal fusion, which may limit their ability to capture complex drug-cell interactions. To address this issue, we developed GBFN, a gated bimodal fusion network for continuous IC50 prediction that integrates pretrained drug and cell-line representations. Specifically, drug embeddings were obtained from SMI-TED, whereas cell-line embeddings were derived from transcriptomic profiles using BulkFormer. These two modalities were then combined through a dimension-wise gated fusion module and used to predict IC50 values in matched drug-cell line pairs. On the CCLE-based benchmark, GBFN outperformed representative neural baselines, including GraphDRP, TGSA, and TransEDRP, and achieved the best overall performance, with an R² of 0.8714 and an RMSE of 0.8938. Moreover, ablation analysis showed that the model using drug features and cell-line expression data with gated fusion performed better than the corresponding model using direct concatenation, indicating that the improvement was associated with the fusion strategy rather than with the input modalities alone. In addition, cell-line expression data were more informative than mutation data in the present setting, and adding mutation data to the model using drug features and expression data did not further improve performance. Across major cancer types, GBFN maintained generally high cell-line-level predictive performance, and perturbation-based attribution identified biologically relevant transcriptomic programs in selected drug-cell line settings. Together, these findings support GBFN as a compact and effective framework for continuous drug response prediction.

Humans↗

PubMed2026

GICPIdb: an archival repository of multimodal data focusing on pathological images for gastrointestinal cancers.

INTRODUCTION: Deep learning (DL) shows great potential for predicting biomarkers from routine histopathological slides of gastrointestinal (GI) cancers. Yet most existing models are validated on limited patient cohorts, while pathological image annotation and molecular marker standardization demand substantial professional expertise. To address these gaps, we constructed the Gastrointestinal Cancer Pathological Image Archive (GICPIdb, gicpidb.shubuzuo.top), a dedicated database and web platform covering seven major GI cancer types. METHODS: High-quality hematoxylin and eosin (H&E)-stained whole-slide images were collected from multiple sources and uniformly processed. Image annotations were performed by board-certified pathologists following standardized protocols. GICPIdb offers five interactive web modules for data uploading, quality control, feature extraction, online annotation and AI-based prediction. Its intuitive interface supports data browsing, retrieval, visualization and downloading. RESULTS: The database houses 2,863 pathologist-annotated, uniformly processed, high-quality H&E stained images collected from 2,655 patients. Of these, 1,699 patients were sourced from The Cancer Genome Atlas (TCGA), 182 from the Clinical Proteomic Tumor Analysis Consortium (CPTAC), and 424 from China-Japan Friendship Hospital and 350 from Chifeng Municipal Hospital in Inner Mongolia, China. It also integrates data on over 50 key molecular markers (e.g., MSI, TMB) and prognostic labels related to survival, recurrence and metastasis. DISCUSSION: GICPIdb aims to promote the development of DL-driven AI tools for cancer research and clinical translation. The multi-institutional data collection and standardized annotation pipeline are expected to enhance the generalizability and reproducibility of AI-based prediction models across diverse patient populations.

deep learning↗

PubMed2025

CNV-Finder: Streamlining Copy Number Variation Discovery.

Copy Number Variations (CNVs) play pivotal roles in the etiology of complex diseases and are variable across diverse populations. Understanding the association between CNVs and disease susceptibility is significant in disease genetics research and often requires analysis of large sample sizes. One of the most cost-effective and scalable methods for detecting CNVs is based on normalized signal intensity values, such as Log R Ratio (LRR) and B Allele Frequency (BAF), from Illumina genotyping arrays. In this study, we present CNV-Finder, a novel pipeline integrating deep learning techniques on array data, specifically a Long Short-Term Memory (LSTM) network, to expedite the large-scale identification of CNVs within predefined genomic regions. This facilitates efficient prioritization of samples for time-consuming or costly subsequent analyses such as Multiplex Ligation-dependent Probe Amplification (MLPA), short-read, and long-read whole genome sequencing. We incorporate four genes to establish our methods-Parkin (PRKN), Leucine Rich Repeat And Ig Domain Containing 2 (LINGO2), Microtubule Associated Protein Tau (MAPT), and alpha-Synuclein (SNCA)-which may be relevant to neurological diseases such as Alzheimer's disease (AD), Parkinson's disease (PD), Progressive Supranuclear Palsy (PSP), or related disorders such as essential tremor (ET). By training our models on expert-annotated samples and validating them across diverse cohorts, including those from the Global Parkinson's Genetics Program (GP2) and additional dementia-specific databases, we demonstrate the efficacy of CNV-Finder in accurately detecting deletions and duplications. Our pipeline outputs app-compatible files for visualization within CNV-Finder's interactive web application. This interface enables researchers to review predictions and filter displayed samples by model prediction values, LRR range, and variant count in order to explore or confirm results. Our pipeline integrates this human feedback to enhance model performance and reduce false positive rates. Through a series of comprehensive analyses and validations using visual inspection, MLPA, short-read, and long-read sequencing data, we demonstrate the robustness and adaptability of CNV-Finder in identifying CNVs with regions of varied size, probe density, and noise. Our findings highlight the significance of contextual understanding and human expertise in enhancing the precision of CNV identification, particularly in complex genomic regions like 17q21.31. The CNV-Finder pipeline is a scalable, publicly available resource for the scientific community, available on GitHub (https://github.com/GP2code/CNV-Finder; DOI 10.5281/zenodo.14182563). CNV-Finder not only expedites accurate candidate identification but also significantly reduces the manual workload for researchers, enabling future targeted validation and downstream analyses in regions or phenotypes of interest.

Copy Number Variation (CNV)↗

PubMed2026

Mechanistic insights into Claudin-14 dysfunction implicated in veins of Galen malformation.

Claudin-14 (CLDN14) is a key component of tight junctions (TJs) critical for maintaining paracellular barrier function. Variants of CLDN14 have been linked to Vein of Galen malformations (VOGMs), a rare cerebrovascular disorder; however, the molecular mechanisms underlying their pathogenicity remain unknown. Here, we investigate the mechanistic effects of two VOGM-associated mutations, A113P and V143M, using reinforcement-learning driven enhanced sampling molecular dynamics simulations combined with DiffNets-based deep learning and independent trajectory-wide structural analyses. Our analysis reveals that A113P induces broader structural disruption of CLDN14, perturbing paracellular sealing, pore symmetry, and inter-protomer communication, whereas V143M induces structural rearrangements centred around TM3 and the TM3-ECL2 region. In both cases, mutation-specific alterations are observed in structural stability and interfacial organization across oligomeric assemblies. Notably, these effects are qualitatively consistent across different modelled architectures, despite variability in local responses. In the absence of experimentally resolved structures, the structural perturbations reported here provide a mechanistic understanding of how VOGM-associated variants may influence CLDN14 structure and dynamics.

Aneurysm↗

PubMed2026

Cancer of unknown primary: the evolution of tissue of origin identification in the artificial intelligence era.

Cancer of Unknown Primary (CUP) presents substantial diagnostic and therapeutic challenges owing to its heterogeneous nature and the absence of an identifiable primary tumor site. This review provides a structured search of the pathogenesis, epidemiological characteristics, and limitations of traditional diagnostic and therapeutic approaches for CUP, with an emphasis on the evolution of Tissue of Origin (TOO) identification techniques. Recent advances in precision medicine have accelerated the development of machine learning-based TOO identification tools, representing a paradigm shift in CUP diagnostics. Deep learning (DL) algorithms that integrate multi-omics data (such as genomics and transcriptomics) with clinical features have markedly enhanced the accuracy of tracing tumor origin, and artificial intelligence (AI) driven TOO models are increasingly being incorporated into clinical practice, offering new insights for pathological diagnosis, treatment selection, and prognostic evaluation. Nevertheless, several challenges remain, including issues of data standardization, model generalizability, and interpretability. Ethical considerations related to data privacy, algorithmic fairness, and clinical implementation also warrant careful attention. Future research should focus on establishing standardized multi-center databases, developing more interpretable AI models, and fostering multidisciplinary collaborative strategies for CUP management. Through continued refinement of technical solutions and regulatory guidelines, TOO identification is anticipated to progress from research to routine clinical application, ultimately supporting precise and personalized care for patients with CUP.

Artificial intelligence↗

PubMed2026

Development and Validation of Machine Learning Models for Predicting Early Cognitive Decline Using Home Sensor-Derived Behavioral Data: Sensors in-Home for Elder Wellbeing (SINEW) Cohort Study.

BACKGROUND: As the global population continues to age, the prevalence of geriatric conditions, including dementia and frailty, is also increasing. Early identification of individuals at an elevated risk of these conditions, such as those presenting with mild cognitive impairment (MCI) or prefrailty, can provide a critical window for prompt intervention aimed at preventing or reversing disease progression. To promote such early identification, there is a burgeoning interest in the use of digital sensor technology and predictive modeling. OBJECTIVE: This study aimed to use a continuous, home-based monitoring sensor system for older adults to distinguish those exhibiting normal aging from those with MCI, early dementia, prefrailty, or frailty, and to predict their transition from normal aging to one of these conditions. METHODS: This longitudinal cohort study will recruit 200 community-dwelling adults aged ≥65 years with normal cognition or MCI at baseline. A multi-sensor system will be installed in participants' homes, including passive infrared motion sensors, door contact sensors, bed sensors, medication box sensors, wearable activity bands, and Bluetooth proximity beacons. These devices will continuously capture spatiotemporal activity patterns, mobility indicators, sleep behaviors, and medication-taking routines. Annual assessments will include standardized cognitive tests (eg, Montreal Cognitive Assessment, Mini-Mental State Examination, Rey Auditory-Verbal Learning Test, digit span, Color Trails Test, semantic fluency, Stroop), frailty measures (modified Fried phenotype, gait speed, grip strength), mental health scales, sleep quality, and psychosocial indicators. Sensor-derived features-such as gait variability, activity regularity, sleep fragmentation, and medication adherence patterns-will be integrated with clinical data to develop supervised machine learning models. Planned approaches include logistic regression, random forests, gradient boosting, and deep learning. Model performance will be evaluated using cross-validation and independent test sets. Primary metrics will include area under the receiver operating characteristic curve, sensitivity, specificity, precision, recall, and F1-score. Models will be benchmarked against gold-standard clinical diagnoses and validated using temporal subsets of the dataset. RESULTS: Enrollment for this study started in November 2019 and will continue until March 2030. As of June 2025, we have enrolled 138 participants. Full data analysis has yet to begin. CONCLUSIONS: We aim to develop a reliable and effective sensor system for in-home use that will facilitate the early detection of cognitive and physical decline. In so doing, it will add to our current understanding of digital biomarkers. It is common for older adults to seek clinical intervention only when their cognitive impairment has already reached an advanced stage. The implementation of readily deployable sensor systems within community settings presents us with opportunities for prompt intervention, which holds the potential for delaying or reversing disease progression and allowing for a greater number of functional and meaningful years.

Humans↗

PubMed1996

A relational view of learning: implications for nurse education.

For about two decades, student learning in the context of professional and higher education, more generally, has been investigated by groups of researchers in countries such as Sweden, the UK, Australia and South Africa. The focus in many of these studies has been on the experience of students as they undertake complex, realistic tasks such as reading academic articles, listening to lectures, writing essays, solving problems and learning subject matter concepts. Central to this research is the adoption of a relational and holistic model of learning in which the relationships among all elements in the learning situation, including the student, the learning task, the teaching methods and assessment practices have been investigated. In particular, researchers from this line of inquiry have studied the relationship between the ways students go about learning in natural educational settings, referred to as approaches to learning, and what they learn. They have established that a full understanding of subject matter is reliant on students employing deep learning approaches which, in turn, depend on the perceived learning environment encouraging such learning. In this paper, the work of the relational theorists is described and the substantial and practical findings arising from this school of research are reported. Specific implications for the education of undergraduate nurses are also outlined.

Clinical Competence↗

PubMed2026

ProMeta: a meta-learning framework for robust disease diagnosis and prediction from plasma proteomics.

MOTIVATION: The plasma proteome offers a dynamic window of human health, capturing the real-time intersections between genetics and physiology. However, the application of deep learning to proteomics is currently hindered by a reliance on large-scale labeled datasets, rendering standard models ineffective for rare or novel diseases where patient samples are inherently scarce. RESULTS: Here, we present ProMeta, a meta-learning framework designed to enable robust disease modeling under extreme data restrictions. By integrating knowledge-guided pathway encoding with bi-level meta-optimization, ProMeta projects unstructured proteomic profiles into biologically interpretable functional tokens. This architecture allows the model to learn a global initialization containing transferable biological priors from biobank-scale data, facilitating rapid adaptation to novel tasks. Through comprehensive benchmark experiments, ProMeta consistently outperformed transfer learning and traditional machine learning baselines in both disease diagnosis and prediction tasks. In the most challenging 4-shot scenarios (utilizing only 2 cases and 2 controls), the model achieved robust generalization with an average AUROC of ∼0.69, representing a 24.6% relative improvement over the best-performing baseline methods. Mechanistic investigation revealed that ProMeta disentangles cases from controls in the latent space prior to task-specific adaptation, confirming the acquisition of universal biological rules rather than rote memorization. Furthermore, gradient-based interpretation identified disease-specific protein biomarkers and functional pathways consistent with known pathophysiology. Collectively, ProMeta overcomes the data-scarcity bottleneck in precision medicine, providing a scalable, interpretable framework for characterizing the full spectrum of human diseases, particularly for rare conditions lacking extensive clinical cohorts. AVAILABILITY AND IMPLEMENTATION: The source code of ProMeta is available at GitHub (https://github.com/lihan97/ProMeta).

Proteomics↗

PubMed2026

Diagnostic Performance of Machine Learning for Systemic Lupus Erythematosus: Systematic Review and Meta-Analysis.

BACKGROUND: Early and accurate diagnosis of systemic lupus erythematosus (SLE) and its organ involvement is essential. Previous reviews of machine learning (ML) in SLE combined heterogeneous tasks and validation strategies and may have overinterpreted model performance. OBJECTIVE: This study evaluated the diagnostic performance of ML and deep learning (DL) models for 3 clinically distinct SLE-related tasks: SLE classification or diagnosis, lupus nephritis (LN) diagnosis, and neuropsychiatric systemic lupus erythematosus (NPSLE) discrimination. We also assessed methodological quality and certainty of evidence. METHODS: PubMed, Embase, Cochrane Library, Web of Science, and IEEE Xplore were searched from January 2014 to April 2026. Eligible peer-reviewed diagnostic accuracy studies developed or validated ML or DL models for 1 of the 3 prespecified tasks, used an accepted reference standard, and provided data for a 2×2 contingency table. Bivariate random-effects meta-analyses with the Hartung-Knapp-Sidik-Jonkman adjustment were used to pool sensitivity and specificity. We reported 95% prediction intervals (PIs), assessed risk of bias using the Quality Assessment of Diagnostic Accuracy Studies for Artificial Intelligence tool (QUADAS-AI; Viknesh Sounderajah [Imperial College London]), and evaluated certainty of evidence using the Grading of Recommendations Assessment, Development, and Evaluation framework for diagnostic test accuracy. RESULTS: Twenty-nine studies were included: 17 for SLE classification, 5 for LN diagnosis, and 7 for NPSLE discrimination. In the primary task-stratified analysis, pooled sensitivity was 0.91 (95% CI 0.86-0.94; 95% PI 0.56-0.99), and pooled specificity was 0.94 (95% CI 0.91-0.96; 95% PI 0.69-0.99), with low heterogeneity (I²=23.9% and 22.9%, respectively). DL models showed a sensitivity of 0.93 and specificity of 0.95, compared with 0.88 and 0.94 for traditional ML models. Certainty of evidence was high for most analyses but low for LN diagnosis because of inconsistency and imprecision. All studies were retrospective, and only 9 of 29 (31%) performed independent external validation. Overall risk of bias was high or unclear in 22 of 29 (75.9%) studies. No study reported model calibration, decision-curve analysis, or net clinical benefit. CONCLUSIONS: ML models showed promising diagnostic accuracy across 3 distinct SLE-related tasks, but wide PIs, limited external validation, and pervasive risk of bias restrict conclusions about real-world generalizability. Prospective multicenter studies with standardized tasks and reference standards, independent external validation, and formal assessment of calibration and clinical utility are required before clinical implementation.

Humans↗