Frontiers in Molecular Biosciences最新文献

{"title":"Identification of key genes CCL5, PLG, LOX and C3 in clear cell renal cell carcinoma through integrated bioinformatics analysis.","authors":"Zhenwei Xie, Cheng Feng, Yude Hong, Libo Chen, Mingyong Li, Weiming Deng","doi":"10.3389/fmolb.2025.1587196","DOIUrl":"10.3389/fmolb.2025.1587196","url":null,"abstract":"<p><strong>Background: </strong>Clear Cell Renal Cell Carcinoma (ccRCC) is a malignant tumor with high mortality and recurrence rates and the molecular mechanism of ccRCC genesis remains unclear. In this study, we identified several key genes associated with the prognosis of ccRCC by using integrated bioinformatics.</p><p><strong>Methods: </strong>Two ccRCC expression profiles were downloaded from Gene Expression Omnibus and one dataset was gained from The Cancer Genome Atlas The Robust Rank Aggregation method was used to analyze the three datasets to gain integrated differentially expressed genes The Gene Ontology and KEGG analysis were performed to explore the potential functions of DEGs. The Search Tool for the Retreival of Interacting Genes/Proteins (STRING) and Cytoscape software were used to construct protein-protein interaction network and module analyses to screen the hub genes. Spearman's correlation analysis was conducted to evaluate the interrelationships among the hub genes. The prognostic value was evaluated through K-M survival analysis, Cox regression analysis, and receiver operating characteristic curve analysis to determine their potential as prognostic biomarkers in ccRCC. The expression of hub genes between ccRCC and adjacent normal tissues was analyzed by RT-qPCR, Western blotting, and immunohistochemical (IHC).</p><p><strong>Result: </strong>125 DEGs were identified using the limma package and RRA method, including 62 up-expressed genes and 63 down-expressed genes. GO and KEGG analysis showed some associated pathways. Spearman's correlation analysis revealed that the hub genes are not only interrelated but also closely associated with immune cell infiltration. Gene expression analysis of the hub genes based on the TCGA-KIRC cohort, along with K-M survival analysis, Cox regression, and ROC curve analysis, consistently demonstrated that CCL5, LOX, and C3 are significantly upregulated in ccRCC and are associated with poor clinical outcomes. In contrast, PLG showed opposite result. These results were further validated at the mRNA and protein levels.</p><p><strong>Conclusion: </strong>Our findings indicate that CCL5, LOX, C3, and PLG are significantly associated with the progression and prognosis of ccRCC, highlighting their potential as prognostic biomarkers. These results provide a foundation for future research aimed at uncovering the underlying mechanisms and identifying potential therapeutic targets for ccRCC.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1587196"},"PeriodicalIF":3.9,"publicationDate":"2025-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12088980/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144110234","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":"Epigenetic modifications in cardiac fibrosis: recent evidence of new pharmacological targets.","authors":"Marian Pérez, Mónica Gómez, Jairo Castellar-López, Patricio Araos, Evelyn Mendoza-Torres, Samir Bolívar","doi":"10.3389/fmolb.2025.1583446","DOIUrl":"10.3389/fmolb.2025.1583446","url":null,"abstract":"<p><p>Cardiac fibrosis (CF) is characterized by the excessive deposition of collagen types I (COI I) and III (COI III), primarily mediated by cardiac fibroblasts (CFB). Recent advances in epigenetic research have enhanced our understanding of the molecular mechanisms underlying CF and have facilitated the identification of novel therapeutic strategies targeting key proteins and signaling pathways involved in its progression. Epigenetic modifications, including DNA methylation, histone modifications, and non-coding RNAs (ncRNAs), are structural and chemical alterations that regulate gene expression and cellular responses without changing the DNA sequence. Investigating the role of epigenetic enzymes in CF may reveal promising pharmacological targets. This review summarizes current evidence on epigenetic modifications implicated in CF and discusses their potential as therapeutic targets for modulating this pathological process.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1583446"},"PeriodicalIF":3.9,"publicationDate":"2025-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12081422/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144093328","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":"Predictive value of CPS combined with inflammatory markers for pathological remission of locally advanced head and neck squamous cell carcinoma after adjuvant immunochemotherapy.","authors":"Yudong Ning, Han Li, Yixuan Song, Yuqin He, Shaoyan Liu, Yang Liu","doi":"10.3389/fmolb.2025.1593742","DOIUrl":"https://doi.org/10.3389/fmolb.2025.1593742","url":null,"abstract":"<p><strong>Objective: </strong>To explore the predictive value of the combined positive score (CPS) and the neutrophil-to- platelet count ratio (NPR) for surgical pathological remission in patients with locally advanced head and neck squamous cell carcinoma (LAHNSCC) who have undergone neoadjuvant immunotherapy combined with chemotherapy (NICC).</p><p><strong>Method: </strong>Patients with LAHNSCC who underwent NICC and surgery from May 2021 to September 2023 were retrospectively analyzed. CPS, NPR and other clinically relevant parameters were collected, which includes gender, age, tumor types, multiple cancer, differentiation, T staging, N staging, immunotherapy cycles and postoperative pathological remission degree.</p><p><strong>Result: </strong>Patients with a higher CPS were significantly associated with a higher pathological complete response (PCR) of the primary site (PPCR) (P = 0.034) and a higher PCR of the lymph nodes (LPCR) (P = 0.085). Specifically, patients with a CPS of ≥20 demonstrated a higher rate of severe pathologic tumor response (PTR), with values of 80.8% compared to 66.7% and 50%. Notably, even patients with a CPS <1 had a relatively high severe PTR rate of 66.7%. Moreover, patients with NPR <0.024 exhibited a higher severe PTR, regardless of the CPS subgroups (P < 0.05).</p><p><strong>Conclusion: </strong>Higher CPS can be considered a good predictor of higher PCR after NICC in patients with LAHNSCC. Patients with CPS <1 can still achieve a higher PTR. Patients with NPR <0.024 can help achieve a higher severe PTR in patients with LAHNSCC regardless of the CPS.CPS combined with NPR may have a better predicted value for surgical PTR of HNSCC after NICC.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1593742"},"PeriodicalIF":3.9,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12078134/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144075930","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":"Integrative multi-omics and Mendelian randomization analysis reveal SPP1⁺ tumor-associated macrophage-driven prognostic signature for hepatocellular carcinoma.","authors":"Kai Lei, Yichun Lei, Zeyao Wang, Zhixin Ye, Jiawei Liu, Wenhao Chen, Caihong Zhou, Jinmei Tan, Shuxian Chen, Yifan Zhang, Jiehui Tan","doi":"10.3389/fmolb.2025.1594610","DOIUrl":"https://doi.org/10.3389/fmolb.2025.1594610","url":null,"abstract":"<p><strong>Background: </strong>The SPP1⁺ tumor-associated macrophages (TAMs) have been implicated in tumor metastasis and immune evasion. However, the prognostic significance of SPP1⁺ TAMs in hepatocellular carcinoma (HCC) remains largely unexplored. This study aimed to identify SPP1⁺ TAMs-related genes and construct a model to predict overall survival (OS) in HCC patients.</p><p><strong>Methods: </strong>Single-cell RNA sequencing (scRNA-seq) datasets from HCC patients were analyzed to identify SPP1⁺ TAMs. SPP1⁺ TAMs-related risk score (STRS) was developed using Mendelian randomization (MR) analysis and Least Absolute Shrinkage and Selection Operator (LASSO) regression. HCC patients from the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) cohorts were stratified into high- and low-STRS groups based on STRS. Kaplan-Meier survival analysis, receiver operating characteristic (ROC) curve analysis, and functional enrichment analysis were performed to assess the prognostic value of STRS.</p><p><strong>Results: </strong>SPP1⁺ TAMs exhibited strong associations with immunosuppressive functions. 16 SPP1⁺ TAMs-related genes were used to construct STRS. Patients in the high-STRS group had significantly worse OS than those in the low-STRS group (<i>p</i> < 0.001). ROC analysis demonstrated robust predictive power, with AUC values ranging from 0.685 to 0.748 for 1-year OS, 0.717 to 0.739 for 2-year OS, and 0.719 to 0.738 for 3-year OS. The STRS model also exhibited strong predictive capability for the distinction of drug resistance.</p><p><strong>Conclusion: </strong>This study identified SPP1⁺ TAMs-related genes as key prognostic indicators in HCC. The STRS model provides an effective tool for predicting patient survival and may facilitate personalized treatment strategies for HCC. These findings enhance the understanding of TAMs-driven immune modulation in HCC and highlight potential therapeutic targets for improving patient outcomes.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1594610"},"PeriodicalIF":3.9,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12078150/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144077157","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":"Editorial: Structural studies of bacterial and viruses.","authors":"Annalisa Pastore","doi":"10.3389/fmolb.2025.1605596","DOIUrl":"https://doi.org/10.3389/fmolb.2025.1605596","url":null,"abstract":"","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1605596"},"PeriodicalIF":3.9,"publicationDate":"2025-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12075936/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144077117","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":"Glycosyltransferases: glycoengineers in human milk oligosaccharide synthesis and manufacturing.","authors":"Alanna S Slater, Andrew G McDonald, Rita M Hickey, Gavin P Davey","doi":"10.3389/fmolb.2025.1587602","DOIUrl":"https://doi.org/10.3389/fmolb.2025.1587602","url":null,"abstract":"<p><p>Human milk oligosaccharides (HMOs) are a diverse group of complex carbohydrates that play crucial roles in infant health, promoting a beneficial gut microbiota, modulating immune responses, and protecting against pathogens. Central to the synthesis of HMOs are glycosyltransferases, a specialized class of enzymes that catalyse the transfer of sugar moieties to form the complex glycan structures characteristic of HMOs. This review provides an in-depth analysis of glycosyltransferases, beginning with their classification based on structural and functional characteristics. The catalytic activity of these enzymes is explored, highlighting the mechanisms by which they facilitate the precise addition of monosaccharides in HMO biosynthesis. Structural insights into glycosyltransferases are also discussed, shedding light on how their conformational features enable specific glycosidic bond formations. This review maps out the key biosynthetic pathways involved in HMO production, including the synthesis of lactose, and subsequent fucosylation and sialylation processes, all of which are intricately regulated by glycosyltransferases. Industrial methods for HMO synthesis, including chemical, enzymatic, and microbial approaches, are examined, emphasizing the role of glycosyltransferases in these processes. Finally, the review discusses future directions in glycosyltransferase research, particularly in enhancing the efficiency of HMO synthesis and developing advanced analytical techniques to better understand the structural complexity and biological functions of HMOs.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1587602"},"PeriodicalIF":3.9,"publicationDate":"2025-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12074965/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144077082","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":"<i>Cuscutae Semen</i> in depression-induced ovarian dysfunction: metabolomics with UPLC-QToF-MS in female mice.","authors":"Ying Xie, Zhaoxiang Zeng, Jinrong Zhang, Qiangqiang Han, Chengwu Song, Shuna Jin, Min Zhao","doi":"10.3389/fmolb.2025.1595602","DOIUrl":"https://doi.org/10.3389/fmolb.2025.1595602","url":null,"abstract":"<p><p>The increasing prevalence of depression profoundly affects female ovarian health. Although <i>Cuscutae Semen</i> (CS) is acknowledged for treating reproductive disorders, its pharmacological mechanisms in depression-induced ovarian dysfunction remain insufficiently explored. This study investigated CS's effects in a chronic unpredictable mild stress (CUMS) mouse model of depression. Mice were divided into control, CUMS model, CS treatment and estradiol treatment group. Behavioral and biochemical analyses assessed depressive-like behaviors and hormone levels. Untargeted metabolomics utilizing ultra-performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry was applied to identify differential metabolites of CS in the treatment of depression-induced ovarian dysfunction. These findings were confirmed through real-time quantitative polymerase chain reaction assays. Based on the outcomes from behavioral and biochemical assays, CS effectively ameliorated the chronic unpredictable mild stress-induced reproductive ailment in mice. Ten differential metabolites were identified, highlighting the impact of CUMS and CS's ameliorative effects. Pathways linked to arachidonic acid metabolism, glycerophospholipid metabolism, linoleic acid metabolism, and steroid hormone biosynthesis were involved. Seven target genes further validated the metabolomic analysis. This study provides strong evidence of CS's therapeutic potential in alleviating depression-induced ovarian dysfunction, shedding light on its pharmacological mechanisms and supporting its use as a functional medical food.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1595602"},"PeriodicalIF":3.9,"publicationDate":"2025-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12074923/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144077114","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":"Unraveling the role of histone acetylation in sepsis biomarker discovery.","authors":"Feng Cheng, Juxin Deng, Zhaoyang Du, Lei Li, Zhaolei Qiu, Min Zhu, Hongchang Zhao, Zhenjie Wang","doi":"10.3389/fmolb.2025.1582181","DOIUrl":"https://doi.org/10.3389/fmolb.2025.1582181","url":null,"abstract":"<p><strong>Introduction: </strong>Sepsis is a life-threatening condition caused by a dysregulated immune response to infection. Despite advances in clinical care, effective biomarkers for early diagnosis and prognosis remain lacking. Emerging evidence suggests that histone acetylation plays a crucial role in the pathophysiology of sepsis.</p><p><strong>Methods: </strong>Transcriptomic and single-cell RNA sequencing data were used to identify histone acetylation-related genes. Differential expression analysis and weighted gene co-expression network analysis (WGCNA) were performed, followed by machine learning algorithms (LASSO, SVM-RFE, and Boruta) to screen for potential biomarkers. Mendelian randomization (MR), RT-qPCR, and functional assays were conducted for validation.</p><p><strong>Results: </strong><i>BLOC1S1</i>, <i>NDUFA1</i>, and <i>SFT2D1</i> were identified as key biomarkers. A predictive nomogram demonstrated strong diagnostic potential. Immune infiltration and single-cell analyses linked the biomarkers to macrophage activity. MR analysis confirmed <i>SFT2D1</i> as a causal factor in sepsis. Functional assays showed that knockdown of <i>SFT2D1</i> suppressed <i>CXCL10</i> and <i>IL-6</i> expression, indicating its pro-inflammatory role.</p><p><strong>Discussion: </strong>This study identifies novel biomarkers associated with histone acetylation and immune dysregulation in sepsis. These findings deepen our understanding of sepsis pathogenesis and may facilitate the development of improved diagnostic and therapeutic strategies.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1582181"},"PeriodicalIF":3.9,"publicationDate":"2025-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12074977/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144076993","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":"Type-2 diabetes epigenetic biomarkers: present status and future directions for global and Indigenous health.","authors":"Sarah Munns, Alex Brown, Sam Buckberry","doi":"10.3389/fmolb.2025.1502640","DOIUrl":"https://doi.org/10.3389/fmolb.2025.1502640","url":null,"abstract":"<p><p>Type-2 diabetes is a systemic condition with rising global prevalence, disproportionately affecting Indigenous communities worldwide. Recent advances in epigenomics methods, particularly in DNA methylation detection, have enabled the discovery of associations between epigenetic changes and Type-2 diabetes. In this review, we summarise DNA methylation profiling methods, and discuss how these technologies can facilitate the discovery of epigenomic biomarkers for Type-2 diabetes. We critically evaluate previous DNA methylation biomarker studies, particularly those using microarray platforms, and advocate for a shift towards sequencing-based approaches to improve genome-wide coverage. Furthermore, we emphasise the need for biomarker studies that include genetically diverse populations, especially Indigenous communities who are significantly impacted by Type-2 diabetes. We discuss research approaches and ethical considerations that can better facilitate Type-2 diabetes biomarker development to ensure that future genomics-based precision medicine efforts deliver equitable health outcomes. We propose that by addressing these gaps, future research can better capture the genetic and environmental complexities of Type-2 diabetes among populations at disproportionate levels of risk, ultimately leading to more effective diagnostic and therapeutic strategies.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1502640"},"PeriodicalIF":3.9,"publicationDate":"2025-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12066322/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143977430","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":"MRI-based deep learning with clinical and imaging features to differentiate medulloblastoma and ependymoma in children.","authors":"Yasen Yimit, Parhat Yasin, Yue Hao, Abudouresuli Tuersun, Chencui Huang, Xiaoguang Zou, Ya Qiu, Yunling Wang, Mayidili Nijiati","doi":"10.3389/fmolb.2025.1570860","DOIUrl":"https://doi.org/10.3389/fmolb.2025.1570860","url":null,"abstract":"<p><strong>Background: </strong>Medulloblastoma (MB) and ependymoma (EM) in children share similarities in terms of age group, tumor location, and clinical presentation, which makes it challenging to clinically diagnose and distinguish them.</p><p><strong>Purpose: </strong>The present study aims to explore the effectiveness of T2-weighted magnetic resonance imaging (MRI)-based deep learning (DL) combined with clinical imaging features for differentiating MB from EM.</p><p><strong>Methods: </strong>Axial T2-weighted MRI sequences obtained from 201 patients across three study centers were used for model training and testing. The regions of interest were manually delineated by an experienced neuroradiologist with supervision by a senior radiologist. We developed a DL classifier using a pretrained AlexNet architecture that was fine-tuned on our dataset. To mitigate class imbalance, we implemented data augmentation and employed K-fold cross-validation to enhance model generalizability. For patient classification, we used two voting strategies: hard voting strategy in which the majority prediction was selected from individual image slices; soft voting strategy in which the prediction scores were averaged across slices with a threshold of 0.5. Additionally, a multimodality fusion model was constructed by integrating the DL classifier with clinical and imaging features. The model performance was assessed using a 7:3 random split of the dataset for training and validation, respectively. The key metrics like sensitivity, specificity, positive predictive value, negative predictive value, F1 score, area under the receiver operating characteristic curve (AUC), and accuracy were calculated, and statistical comparisons were performed using the DeLong test. Thereafter, MB was classified as positive, while EM was classified as negative.</p><p><strong>Results: </strong>The DL model with the hard voting strategy achieved AUC values of 0.712 (95% confidence interval (CI): 0.625-0.797) on the training set and 0.689 (95% CI: 0.554-0.826) on the test set. In contrast, the multimodality fusion model demonstrated superior performance with AUC values of 0.987 (95% CI: 0.974-0.996) on the training set and 0.889 (95% CI: 0.803-0.949) on the test set. The DeLong test indicated a statistically significant improvement in AUC values for the fusion model compared to the DL model (<i>p</i> < 0.001), highlighting its enhanced discriminative ability.</p><p><strong>Conclusion: </strong>T2-weighted MRI-based DL combined with multimodal clinical and imaging features can be used to effectively differentiate MB from EM in children. Thus, the structure of the decision tree in the decision tree classifier is expected to greatly assist clinicians in daily practice.</p>","PeriodicalId":12465,"journal":{"name":"Frontiers in Molecular Biosciences","volume":"12 ","pages":"1570860"},"PeriodicalIF":3.9,"publicationDate":"2025-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12066621/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143993514","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}