Data Sheet 2_Comprehensive profiling of chemokine and NETosis-associated genes in sarcopenia: construction of a machine learning-based diagnostic nomogram.csv

BackgroundChemokines and neutrophil extracellular trap formation (NETosis) are critical drivers of inflammatory responses. However, the molecular characteristics and interaction mechanisms of these processes in sarcopenia remain incompletely understood.
MethodsUtilizing the mRNA expression profile dataset GSE226151 (including 19 sarcopenia, 19 pre-sarcopenia, and 20 healthy control samples), enrichment analysis was performed to identify differentially expressed NETosis-related genes (DENRGs) and chemokine-related genes (DECRGs). Two machine learning algorithms and univariate analysis were integrated to screen signature genes, which were subsequently used to construct diagnostic nomogram models for sarcopenia. Single-gene Gene Set Enrichment Analysis (GSEA) and Gene Set Variation Analysis (GSVA) were used to investigate pathway associations, followed by the construction of a gene interaction network.
ResultsA total of 7 DECRGs and DENRGs were identified, primarily enriched in chemokine signaling pathways, cytokine-cytokine receptor interactions, and sarcopenia-related diseases. Machine learning and univariate analysis revealed three signature genes (CXCR1, CXCR2, and LPL). The nomogram models demonstrated high predictive accuracy in distinguishing sarcopenia from both healthy and pre-sarcopenic states, as evidenced by AUC values of 0.837 (95% CI 0.703–0.947) and 0.903 (95% CI 0.789–0.989), respectively. Single-gene GSEA highlighted significant associations between these genes and the JAK-STAT and PPAR signaling pathways. GSVA indicated that sarcopenia was closely linked to upregulated chemokine signaling, cytokine-receptor interaction activities, and leukocyte transendothelial migration.
ConclusionThe research pinpointed three genes associated with chemokines and NETosis (CXCR1, CXCR2, LPL) and developed highly accurate diagnostic models, offering a new and preliminary approach to differentiate sarcopenia and its early stages.

研究背景：趋化因子（Chemokines）与中性粒细胞胞外陷阱形成（NETosis）是炎症反应的关键驱动因素。然而，肌少症中这两类过程的分子特征及相互作用机制目前仍未完全阐明。研究方法：本研究采用mRNA表达谱数据集GSE226151（包含19例肌少症样本、19例肌少症前期样本及20例健康对照样本），通过富集分析筛选差异表达的NETosis相关基因（DENRGs）与趋化因子相关基因（DECRGs）。整合两种机器学习算法与单变量分析以筛选特征基因，随后基于这些特征基因构建肌少症诊断列线图模型。采用单基因基因集富集分析（GSEA）与基因集变异分析（GSVA）探究通路关联，并进一步构建基因相互作用网络。研究结果：本研究共鉴定得到7个DECRGs与DENRGs，这些基因主要富集于趋化因子信号通路、细胞因子-细胞因子受体相互作用通路及肌少症相关疾病通路。通过机器学习与单变量分析，最终确定3个特征基因：CXCR1、CXCR2及LPL。列线图模型在区分肌少症与健康状态、肌少症前期状态时均表现出较高的预测效能，对应的受试者工作特征曲线下面积（AUC）值分别为0.837（95%置信区间：0.703~0.947）与0.903（95%置信区间：0.789~0.989）。单基因GSEA分析显示，上述特征基因与JAK-STAT及PPAR信号通路存在显著关联。GSVA分析表明，肌少症与趋化因子信号通路、细胞因子-受体相互作用活性及白细胞跨内皮迁移的上调密切相关。研究结论：本研究明确了3个与趋化因子及NETosis相关的基因（CXCR1、CXCR2、LPL），并构建了高精度的诊断模型，为肌少症及其早期阶段的鉴别提供了全新的初步研究思路。