Biodata Mining最新文献

{"title":"Automatic extraction of mesenchymal stromal cells from single-cell RNA-sequencing data of human dental pulp.","authors":"Daigo Okada, Tomoko Takeda-Kawaguchi, Ken-Ichi Tezuka","doi":"10.1186/s13040-026-00581-x","DOIUrl":"10.1186/s13040-026-00581-x","url":null,"abstract":"<p><strong>Background: </strong>Human dental pulp is a complex tissue composed of diverse cell types, including mesenchymal stromal cells (MSCs), which are crucial for tissue repair and regeneration. Although single-cell RNA sequencing (scRNA-seq) data from human dental pulp have accumulated in recent years, MSC identification relies on manual verification of marker genes after clustering, which limits analytical efficiency and scalability. The choice of clustering resolution is determined empirically, potentially leading to under- or over-segmentation of MSCs or their mixing with other cell types. The objective of this study was to establish a computational workflow that automatically extracted uncultured MSCs from human dental pulp scRNA-seq data and to investigate the characteristics of freshly-isolated MSCs in dental pulp using multiple public datasets.</p><p><strong>Results: </strong>A computational workflow was developed that automatically identified MSC populations given a predefined marker set from a scRNA-seq count matrix and systematically evaluated the performance of the marker set. The MSC marker set consisting of six genes (NT5E, THY1, ENG, FRZB, NOTCH3, MCAM) demonstrated higher cluster separation than the three basic MSC markers (NT5E, THY1, ENG) and consistently detected an MSC population across multiple independent dental pulp scRNA-seq datasets. Pseudo-bulk transcriptomic analysis of MSC populations extracted using this workflow indicated that MSCs in the dental pulp of patients with pulpitis themselves exhibited an inflammatory phenotype, characterized by substantial activation of inflammation-related pathways. Donor aging was associated primarily with reduced mesenchymal identity and metabolic alterations.</p><p><strong>Conclusions: </strong>This study provides a framework for the automated extraction of dental pulp-derived MSCs and cross-dataset analysis. This framework is broadly applicable to the automated extraction of cell populations, such as dental pulp-derived MSCs, for which the annotation method is not yet well established, and is expected to be widely useful in single-cell analyses.</p>","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13435457/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148406498","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Interpretable machine learning for Parkinson's disease diagnosis, staging, and biological mechanism exploration: a multicenter analysis.","authors":"Lu Hao, Zihao Wang, Yutong Zhu, Xizheng Wang, Minghui Zhu, Kalibunuer Mahemuti, Yangtai Guan","doi":"10.1186/s13040-026-00574-w","DOIUrl":"10.1186/s13040-026-00574-w","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;To develop and validate an interpretable machine learning model based on multicenter T1-weighted MRI radiomics data for the three-way classification of Parkinson's disease (PD)-Normal (NM), early-stage, and mid-late-stage-and to clarify the diagnostic value of key cortical nuclear features, with further elucidation of the molecular basis through proteomic association analysis.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;A total of 200 patients from multiple centers were included and divided into a training set (n = 76), an internal validation set (n = 33), and an external validation set (n = 91). Six core nuclei, including the caudate nucleus (CN), putamen (PUT), globus pallidus (GP), red nucleus (RN), substantia nigra (SN), and nucleus accumbens (NAc), were segmented using the DKT template. A total of 107 radiomics features were extracted per nucleus. After a three-step dimension reduction (variance thresholding, univariate selection, and multi-class LASSO), key features were identified. Five machine learning models were constructed. Model performance was evaluated using micro/macro-AUC and accuracy (ACC). SHAP analysis was employed to reveal the contribution of specific features to the diagnostic process. Additionally, proteomic analysis was performed on a subset of patients from Center 1 to screen for differentially expressed proteins among groups, and correlation analysis between DEPs and key radiomic features was conducted to explore biological mechanisms.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Nine key features were finally selected, predominantly consisting of gray-level non-uniformity within the PUT. The GBT model performed optimally, with macro-AUC/micro-AUC of 0.890/0.902 in the internal validation set and 0.876/0.869 in the external validation set, respectively. SHAP analysis indicated that the NM group primarily relied on texture features of the PUT and NAc; early-stage PD was characterized by structural changes in the GP and PUT; and mid-late-stage PD was correlated with features from the SN and RN. Proteomic analysis identified 514 DEPs between early PD and controls (411 upregulated, 103 downregulated), which were enriched in cytoskeletal and immune pathways (FDR &lt; 0.05), and 123 DEPs between mid-to-late and early PD (26 upregulated, 97 downregulated), which were enriched in proteasome and neurodegeneration pathways (FDR &lt; 0.05). Multiple significant correlations between key radiomic features and DEPs were identified (all FDR-corrected P &lt; 0.01), including original_gldm_DependenceNonUniformity_PUT with HGFAC (r = 0.510), original_ngtdm_Contrast_GP with PLAT (r = 0.453), and original_glrlm_RunEntropy_GP_QSM with both ASPN (r = 0.62) and TFPI (r = - 0.58).&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusion: &lt;/strong&gt;Interpretable machine learning models based on multicenter T1-weighted nuclear radiomics features can accurately diagnose and stage PD. Integration of radiomics and proteomics enhances model interpretability by linking imaging features ","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13579904/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148340702","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Learning a distance for the clustering of patients with amyotrophic lateral sclerosis.","authors":"Guillaume Tejedor Y Dominguez, Hugo Alarcan, Veronika Peralta, Nicolas Labroche, Philippe Corcia, Patrick Marcel, Hélène Blasco","doi":"10.1186/s13040-026-00579-5","DOIUrl":"10.1186/s13040-026-00579-5","url":null,"abstract":"<p><strong>Background: </strong>Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease with median survival of 3-5 years. Patient responses to treatments vary widely, highlighting the need for personalized care. Clustering patients based on disease progression could improve prognosis, guide clinical decision-making, and optimize clinical trial design. This study aimed to identify robust ALS patient clusters using ALS Functional Rating Scale-Revised (ALSFRS-R) scores and to determine diagnostic parameters predictive of cluster membership, enabling earlier stratification and targeted management.</p><p><strong>Methods: </strong>Data from the Tours ALS center registry (April 1997-October 2023) were analyzed; after preprocessing, 353 patients monitored every three months between January 2004 and July 2023 with ALSFRS-R, clinical, biological, and demographic data were retained. After preprocessing to handle missing or aberrant data, a weakly supervised approach labeled patient pairs based on their ALSFRS-R sequences. These labels were used to train a classifier to learn a distance for off-the-shelf clustering algorithms. Multiple configurations were tested, varying clustering algorithms, dimensionality reduction method, and number of clusters. Random Forest (RF) model predicted cluster membership from diagnostic parameters. Optimal clustering was selected using silhouette score, validated with Kaplan-Meier survival analysis. Stability and robustness were assessed with the Adjusted Rand Index (ARI) and silhouette score respectively. Predictive performance was evaluated using specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV). Diagnostic parameters associated with clusters were identified using Kruskal-Wallis and chi-squared tests for continuous and categorical variables.</p><p><strong>Results: </strong>Three clusters (n = 139, 121, 93) were identified, demonstrating strong separation (silhouette ≈ 0.6) and high stability of results (ARI ≈ 0.7). Survival differed significantly among clusters: over 50% of patients in the third cluster survived beyond 50 months, compared to less than 25% in the other clusters. Thirteen diagnostic parameters-including ALSFRS-R subscores, IgG levels, albumin quotient, and time to diagnosis-were key predictors of cluster membership. Cluster prediction achieved specificity and NPV ≈ 0.75, with close sensitivity and PPV compared to state-of-the-art methods.</p><p><strong>Conclusion: </strong>This framework successfully stratifies ALS patients into clinically meaningful clusters, revealing underlying disease heterogeneity and providing strong prognostic insight. Such classification can facilitate personalized care, guide therapeutic decisions, and inform the design of targeted interventions to improve outcomes.</p><p><strong>Clinical trial number: </strong>Not applicable.</p>","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13563701/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148333513","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-domain feature fusion with variational mode decomposition and hybrid LightGBM-Logistic Regression for multi-class seizure classification.","authors":"A Aameer Arshath, D Karthikeyan","doi":"10.1186/s13040-026-00578-6","DOIUrl":"10.1186/s13040-026-00578-6","url":null,"abstract":"<p><strong>Objective: </strong>The problem of seizure classification using multi-seizure types based on reduced-channel EEG recordings is difficult due to the complexity and nonlinearity of seizure patterns and their high class imbalance.</p><p><strong>Methods: </strong>In this paper, a reduced-channel EEG-based method is proposed to perform automated multi-seizure classification using the Temple University Seizure Corpus (TUSZ) data set. The proposed approach uses the T3 and T4 EEG temporal channels and performs Variational Mode Decomposition (VMD) to decompose EEG signals into several intrinsic mode functions to extract the relevant temporal and spectral features of seizures. Next, multi-domain features, including statistical, spectral, entropy, and nonlinear features, are extracted from each VMD mode. To avoid the redundancy of the extracted features and preserve seizure-related features, an MI-RFE feature selection scheme is used. The extracted features were then classified using a stacked ensemble model incorporating LightGBM and logistic regression. For the purpose of ensuring unbiased testing and avoiding any information leakage, the Fold-Wise SMOTE technique was used in the process of stratified K-fold cross-validation.</p><p><strong>Results: </strong>In experimental analysis of the developed technique with nine types of seizures, positive outcomes were obtained in terms of classification accuracy that included high classification accuracy, high weighted F1 measure, and multiclass ROC-AUC characteristics, resulting in classification accuracy of 92.16%. Also, when comparing the developed feature selection approach with standard techniques, the approach's efficiency was demonstrated.</p><p><strong>Significance: </strong>The attained results show that the suggested approach for reducing the number of channels used in analyzing brain dynamics is effective.</p><p><strong>Clinical trial number: </strong>Not applicable.</p>","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13548600/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148303663","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AVSeg-XAI: Deep learning framework for A/V segmentation with vascular features reveals retinal oculomics as biomarker for cardiovascular disease.","authors":"Syed Abdullah Basit, Tanvir Alam","doi":"10.1186/s13040-026-00573-x","DOIUrl":"10.1186/s13040-026-00573-x","url":null,"abstract":"<p><strong>Background: </strong>Cardiovascular disease (CVD) is the leading cause of mortality across the globe. Conventional risk stratification often depends on invasive biomarkers or generic risk scores that may underperform in key demographics, particularly women and middle-aged adults. Retinal microvasculature provides a non-invasive window into systemic vascular health (oculomics), but its integration with routine clinical vitals for sex-specific risk assessment remains underexplored.</p><p><strong>Methods: </strong>We developed AVSeg-XAI, an interpretable DINOv2-based deep-learning pipeline for (i) artery/vein (A/V) segmentation and (ii) geometric/topological feature extraction from retinal fundus images. We then implemented a multimodal approach by integrating interpretable retinal vascular biomarkers and non-invasive clinical vitals into a Random Forest classifier to stratify CVD risk in a Qatar Biobank (QBB) cohort. Performance was assessed using stratified 10-fold cross-validation with subgroup analyses by sex and age. Biological plausibility was evaluated through virtual phenotyping, which predicts systemic traits based solely on retinal features.</p><p><strong>Results: </strong>For artery/vein segmentation, the self-supervised AVSeg-XAI achieved a mean Dice score of 0.8368 on the benchmark dataset, exceeding the reported LUNet mean Dice score of 0.8327. Retinal feature extraction module yielded 34 refined vascular features encompassing caliber (CRAE, CRVE, AVR), fractal complexity, tortuosity, and branching topology; a parsimonious 7-feature interpretable subset was also defined for baseline comparison. Clinical inputs were restricted to 15 non-invasive variables obtainable without phlebotomy (e.g., age, BMI, blood pressure). The multimodal fusion model achieved an overall AUC of 0.81 (95% CI: 0.78-0.84). Stratified evaluation demonstrated superior performance in females (AUC 0.83, 95% CI: 0.79-0.87) compared with males (AUC 0.79, 95% CI: 0.75-0.83), with peak performance in the 45-55 year subgroup (AUC 0.90, 95% CI: 0.83-0.97).</p><p><strong>Conclusion: </strong>We present an explainable, non-invasive CVD screening approach that integrates retinal imaging biomarkers with basic clinical vitals. We believe this study advances AI-powered oculomics as a proof-of-concept for community health screening, subject to prospective external validation.</p>","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13527959/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148279549","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large-scale transcriptomic data mining using explainable XGBoost and SHAP reveals shared biomarkers and molecular mechanisms between type-2 diabetes and triple-negative breast cancer for drug repurposing.","authors":"Md Feroj Ahmed, Ovi Sharma, Tasfia Noor, Md Abdul Latif, Md Foysal Ahmed, Moumita Bashak, Dewan Mumdood Ahmed, Md Sanoar Hossain, Md Nurul Haque Mollah","doi":"10.1186/s13040-026-00575-9","DOIUrl":"10.1186/s13040-026-00575-9","url":null,"abstract":"<p><p>Type 2 diabetes (T2D) is considered as a risk factor of triple-negative breast cancer (TNBC). So, there is a significant chance of their co-existence. The management of TNBC with T2D becomes more complex than without T2D due to the conflict of therapies, since some T2D drugs may have bad impact on TNBC and vice-versa. Beside this, drug-drug interaction due to the polypharmacy of multiple drugs, during a one-drug one-disease strategy, may create toxicity or side effect to the patients who are suffering from both diseases simultaneously. Therefore, it is required to explore effective unique drugs as the same treatment for both diseases. This study attempted to contribute in this issue. At first, we identified 36 shared differentially expressed genes (sDEGs) that can separate both TNBC and T2D patients from the control group through integrated transcriptomics analysis. Then top-ranked four sDEGs (S100A9, CIRBP, USP10, and PSMD1) were detected as the overlapping dysregulated shared key-genes (sKGs) through the protein-protein interaction (PPI) network analysis and filtering with a machine learning (ML) approach. The gene regulatory network analysis revealed three key transcriptional (TFs proteins) and post-transcriptional (micro-RNAs) factors of sKGs. The enrichment analysis of sKGs with the GO-terms and KEGG-pathways revealed some crucial molecular functions, biological processes, cellular components, and pathways as the key pathogenetic mechanisms for the development and progression of both diseases. Finally, we recommended sKGs-guided two repurposable common candidate drugs (tepotinib and ursodiol) for both diseases by molecular docking, ADME/T analysis and MD simulation studies. Thus, the results of this study could provide useful insights for researchers and medical professionals for improving the diagnostic and therapeutic strategies for the treatment TNBC with T2D as comorbidity.</p>","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13528018/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148279576","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Navigating the uncharted: AI-driven advances in protein structure, dynamics, interactions and ligand interactions for understudied families.","authors":"Shivani Singh, Ruhani Singh, Sunita Sharma","doi":"10.1186/s13040-026-00577-7","DOIUrl":"10.1186/s13040-026-00577-7","url":null,"abstract":"<p><p>The structural and functional characterization of lesser-known protein families remains a major challenge in modern computational biology. Recent breakthroughs in artificial intelligence (AI) and machine learning (ML) have rapidly advanced these fields. This is particularly evident in overcoming the limitations of traditional modelling approaches, such as classical homology modelling and ab initio folding methods. This review traces the evolution of conventional methods to cutting-edge deep learning frameworks, such as AlphaFold2, RoseTTAFold, and transformer-based architectures. We explore how these AI-driven tools achieve near-experimental accuracy in structure prediction, model protein dynamics, and intrinsic disorder. We also addressed computational approaches for protein-protein interactions (PPIs), central to cellular function and interface-targeted drug design, alongside protein-ligand interactions, including novel generative methods. Two representative case studies targeting orphan G-protein-coupled receptors and intrinsically disordered regions demonstrate the transformative potential of these techniques for previously intractable systems. Despite these advances, significant challenges remain, including the need for experimental validation, effective modelling of protein flexibility, and ethical considerations surrounding AI-generated data. We also compare classical and AI-based structural biology pipelines, summarize key tools (e.g. transformers, graph neural networks, and diffusion models), and offer best practice guidelines for computational modelling and data visualization. These developments provide unprecedented insights into the dark proteome regions of the protein universe, enabling the structural illumination of previously uncharacterized and understudied proteins where structure or function was unknown. This review aims to serve as a roadmap for researchers seeking to harness AI innovations to tackle some of the most challenging aspects of proteomics.</p>","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13523371/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148273501","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MPGR-Net: multi-scale patch graph reasoning for weakly supervised pathological image segmentation.","authors":"Xu Zhang, Huaiju Ge, Yanling Wang, Chongqi Wei, Yang Liu, Xinbin Zhang, Cheng Li","doi":"10.1186/s13040-026-00580-y","DOIUrl":"10.1186/s13040-026-00580-y","url":null,"abstract":"<p><p>Weakly supervised pathological image segmentation aims to localize disease regions using coarse annotations (e.g., image-level labels), which significantly reduces annotation cost but introduces challenges such as imprecise localization and noisy predictions. In this paper, we propose multi-scale patch graph reasoning network (MPGR-Net), a novel framework that integrates multi-scale graph reasoning into transformer-based architectures to address these limitations. Specifically, we first employ a Vision Transformer encoder to extract global-aware patch representations. To explicitly model structured spatial dependencies that are often overlooked by standard self-attention, we introduce a multi-scale patch graph reasoning module, where patch tokens are treated as graph nodes and connected via predefined spatial adjacency at multiple scales. This design enables effective aggregation of local and long-range contextual information while maintaining computational efficiency. Furthermore, we incorporate a lightweight cross-attention mechanism to enhance global-to-local feature interaction through channel-wise modulation. Extensive experiments on pathological image segmentation benchmarks demonstrate that MPGR-Net consistently outperforms existing weakly supervised methods, achieving superior segmentation accuracy and robustness. Ablation studies further validate the effectiveness of the proposed graph reasoning module.</p>","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13505138/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148260230","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DAG-VAERL: a novel causal inference method for building causal gene regulatory networks.","authors":"Teng Long, Sachit Satyal, Yong-Fang Kuo, Jean Gao","doi":"10.1186/s13040-026-00571-z","DOIUrl":"10.1186/s13040-026-00571-z","url":null,"abstract":"<p><strong>Background: </strong>Causal discovery methods provide a powerful tool for uncovering the causal relationships between lncRNAs and target genes in gene regulations. In many causal inference and structure learning tasks, learning the Directed Acyclic Graph (DAG) structure from data is a challenging problem. Traditional DAG learning methods often rely on heuristic searches or strict constraints, which fail to effectively handle complex nonlinear relationships and discrete data.</p><p><strong>Methods: </strong>To address this, we propose a novel deep generative model-DAG-VAERL, which combines Graph Neural Networks (GNN) as well as Reinforcement Learning (RL) frameworks and Graph Attention Networks (GAT) module, leveraging Variational Autoencoders (VAE) to learn the DAG structure. DAG-VAERL is capable of modeling complex dependencies between nodes through GNNs and optimizing the graph structure using RL strategies.</p><p><strong>Results: </strong>We conduct extensive experiments on synthetic and real-world datasets, including Alzheimer's disease data, to validate the superiority of DAG-VAERL in structural discovery and parameter estimation. Experimental results demonstrate that DAG-VAERL significantly outperforms traditional methods in structure recovery, especially when dealing with complex data involving nonlinear and discrete variables.</p><p><strong>Conclusions: </strong>This model not only effectively learns the DAG structure from data but also serves as a powerful tool for causal inference and other graph-bard analysis tasks, providing a new approach for related fields.</p>","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13520353/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148266716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Clinically actionable classification tool for childhood and adult atopic dermatitis via explainable AI: a retrospective cross-sectional study optimizing diagnostic accuracy.","authors":"Rong Zhou, Weiming Zhang, Lingyue Zhou, Qin Zhu, Jinqing Zhang, Shiyu Wang, Chang Wang","doi":"10.1186/s13040-026-00565-x","DOIUrl":"10.1186/s13040-026-00565-x","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Despite the critical role of early identification in managing atopic dermatitis (AD), current studies are limited by the absence of adult-specific risk stratification models and imbalanced cohort data, hindering effective risk profiling.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;We aimed to address this deficiency by developing models for AD diagnostic classification through robust machine learning techniques, with a focus on integrating personalized risk factors to improve the identification and characterization of AD in both children and adults.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;The study employed multiracial cross-sectional data from the National Health And Nutrition Examination Survey (NHANES) in which children (age &lt; 20) and adults (age ≥ 20) with self-reported AD endpoints were extracted from 1999 to 2006. Comprehensive class balancing and machine learning techniques were employed to develop a set of AD classification models for different age groups. The models were assessed using five-fold cross-validation for internal validation and internal hold-out validations using a test set composed of raw data. The optimal model for each subgroup, with the best accuracy to classify AD presence, was determined and assessed by ACC, F1-score, MCC, and ROC curve. Key features and their contributions to each classification decision are interpreted by permutation feature importance (PFI) and SHapley Additive explanations (SHAP). We compared and further analyzed demographic and nutritional indicators that have significant associations.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Of 22,635 qualified participants, 1118 (13.47%) children and 1732 (12.80%) adults were identified as having AD. The pronounced class imbalance within the dataset led to baseline models yielding accuracies in the range of merely 0.500-0.503. Following the implementation of class balancing techniques and diverse modeling methods, the refined model, utilizing accessible variables from both children and adults, achieved markedly enhanced performance (ACC: 0.835-0.885, F1-score: 0.834-0.884, AUC: 0.913-0.957). Explainable AI revealed opposing effects of dietary antioxidants between children and adults, highlighting age-specific associated mechanisms.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusion: &lt;/strong&gt;The classification models for both childhood and adult AD have demonstrated robust discriminative ability and potential utility for case identification in both internal and internal hold-out validations. These models address the challenge in accurate, individualized classification for childhood and adult AD, demonstrating strong discriminative potential in our validation cohorts. We have developed a novel AD diagnostic identification system specifically for both children and adults based on readily accessible demographic, clinical, and nutritional features, which enable the characterization of individuals with a high probability of having AD and the pinpointing of their associated features. This ","PeriodicalId":48947,"journal":{"name":"Biodata Mining","volume":" ","pages":""},"PeriodicalIF":7.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13505055/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148260199","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}