基于机器学习法初步探讨二型糖尿病发病机理-国际医学与数据杂志-CSCIED科技核心评价数据库-手机版

基于机器学习法初步探讨二型糖尿病发病机理

A preliminary exploration of the pathogenesis of type 2 diabetes using machine learning methods

ES评分 0 浏览量：104 下载量：0

DOI	10.12208/j.ijmd.20260048
刊名	国际医学与数据杂志 International Journal of Medicine and Data
年，卷(期)	2026, 10(2)
作者	陈文豪^1,2,3, 郑允文^1,2,3,
作者单位	广东省医学大动物模型重点实验室广东江门; 华南生物医药大动物模型研究院广东江门; 五邑大学药学与食品工程学院广东江门;
摘要	本研究旨在整合单细胞转录组学与系统生物学方法，系统解析二型糖尿病（T2D）胰岛的细胞与分子病理特征。通过对GSE221156数据集的高质量单细胞进行整合分析，构建了T2D胰岛的高分辨率转录图谱，发现T2D中β细胞和δ细胞比例显著降低，且β细胞内部功能亚群失衡。通过跨数据集差异分析获得409个T2D β细胞中共同失调的差异表达基因，并利用加权基因共表达网络分析（hdWGCNA）鉴定出与疾病进展紧密相关的核心模块（red与turquoise模块）。结合LASSO回归、XGBoost与SVM-RFE三种机器学习算法，从关键模块中筛选并验证了一个由CALR、IAPP、LAMTOR5、SCGN组成的四基因核心集合。基于此构建的T2D风险预测模型具有良好的判别能力（AUC = 0.732），其可视化列线图工具为风险评估提供了直观依据。本研究从单细胞尺度揭示了T2D胰岛细胞组成、亚群稳态及基因调控网络的多层次紊乱，鉴定出的核心基因模块为理解T2D发病机制及开发新型诊断标志物提供了新线索。
Abstract	This study aims to integrate single-cell transcriptomics with systems biology approaches to systematically elucidate the cellular and molecular pathological characteristics of islets in type 2 diabetes (T2D). Through an integrated analysis of high-quality single-cell data from the GSE221156 dataset, we constructed a high-resolution transcriptional atlas of T2D islets, revealing a significant reduction in the proportion of β-cells and δ-cells in T2D, as well as an imbalance in functional subpopulations within β-cells. Through cross-dataset differential analysis, we identified 409 differentially expressed genes that are commonly dysregulated in T2D β-cells. Using weighted gene co-expression network analysis (hdWGCNA), we identified core modules (the red and turquoise modules) closely associated with disease progression. Combining three machine learning algorithms—LASSO regression, XGBoost, and SVM-RFE—we screened and validated a four-gene core set comprising CALR, IAPP, LAMTOR5, and SCGN from the key modules. The T2D risk prediction model constructed based on this set demonstrated good discriminatory ability (AUC = 0.732), and its visualized contour plot tool provided an intuitive basis for risk assessment. This study reveals, at the single-cell level, multilevel dysregulation in the composition, subpopulation homeostasis, and gene regulatory networks of T2D pancreatic islet cells. The identified core gene modules provide new insights for understanding the pathogenesis of T2D and developing novel diagnostic biomarkers.
关键词	二型糖尿病；单细胞转录组测序；机器学习；生物标志物
KeyWord	Type 2 diabetes; Single-cell transcriptomics; Machine learning; Biomarkers
基金项目
页码	41-49

参考文献
相关文献

[1] MARTíNEZ-LóPEZ J A, LINDQVIST A, LOPEZ-PASCUAL A, et al. Single-cell mRNA-regulation analysis reveals cell type-specific mechanisms of type 2 diabetes
[J]. Nat Commun, 2025, 16(1): 9475.

[2] XIE X, WU C, DAO F, et al. scRiskCell: A single-cell framework for quantifying islet risk cells and their adaptive dynamics in type 2 diabetes
[J]. Imeta, 2025, 4(4): e70060.

[3] BAO K, CUI Z, WANG H, et al. Pseudotime Ordering Single-Cell Transcriptomic of β Cells Pancreatic Islets in Health and Type 2 Diabetes
[J]. Phenomics, 2021, 1(5): 199-210.

[4] XIE X, WU C, YANG Y, et al. Interpretable machine learning-guided single-cell mapping deciphers multi-lineage pancreatic dysregulation in type 2 diabetes
[J]. Cardiovasc Diabetol, 2025, 24(1): 300.

[5] HAO Y, HAO S, ANDERSEN-NISSEN E, et al. Integrated analysis of multimodal single-cell data
[J]. Cell, 2021, 184(13): 3573-87.e29.

[6] MORABITO S, REESE F, RAHIMZADEH N, et al. hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data
[J]. Cell Rep Methods, 2023, 3(6): 100498.

[7] LANGFELDER P, HORVATH S. WGCNA: an R package for weighted correlation network analysis
[J]. BMC Bioinformatics, 2008, 9: 559.

[8] 陈康, 傅强, 卢建强. Vδ2 T细胞在二型糖尿病中的功能及临床研究
[J]. 国际检验医学杂志, 2020, 41(13): 1560-2+7.

[9] WANG J, WEN S, CHEN M, et al. Regulation of endocrine cell alternative splicing revealed by single-cell RNA sequencing in type 2 diabetes pathogenesis
[J]. Commun Biol, 2024, 7(1): 778.

[10] MORABITO S, MIYOSHI E, MICHAEL N, et al. Single-nucleus chromatin accessibility and transcriptomic characterization of Alzheimer's disease
[J]. Nat Genet, 2021, 53(8): 1143-55.

[11] YU Y S, LEE D S, LIM J H, et al. Identifying potential biomarkers for type 2 diabetes in the adipose tissue of older adults via multiple machine learning algorithms
[J]. Sci Rep, 2025, 15(1): 44904.

[12] WU Q, WANG Z, FAN M, et al. Integrative Multi-Omics and Machine Learning Reveal Shared Biomarkers in Type 2 Diabetes and Atherosclerosis
[J]. Int J Mol Sci, 2025, 27(1).

[13] LAWLOR N, GEORGE J, BOLISETTY M, et al. Single-cell transcriptomes identify human islet cell signatures and reveal cell-type-specific expression changes in type 2 diabetes
[J]. Genome Res, 2017, 27(2): 208-22.

[14] XIN Y, KIM J, OKAMOTO H, et al. RNA Sequencing of Single Human Islet Cells Reveals Type 2 Diabetes Genes
[J]. Cell Metab, 2016, 24(4): 608-15.

[15] LI Z, QIAO Z, ZHENG W, et al. Network Cluster Analysis of Protein-Protein Interaction Network-Identified Biomarker for Type 2 Diabetes
[J]. Diabetes Technol Ther, 2015, 17(7): 475-81.

[16] SHARMA A K, KHANDELWAL R, KUMAR M J M, et al. Secretagogin Regulates Insulin Signaling by Direct Insulin Binding
[J]. iScience, 2019, 21: 736-53.

引用本文

陈文豪^,,, 郑允文^,,. 基于机器学习法初步探讨二型糖尿病发病机理 [J]. 国际医学与数据杂志. 2026; 10; (2). 41 - 49.

文献评论

相关学者

相关机构