Department of Joint and Sports Medicine, Weifang Sunshine Union Hospital, Weifang, Shandong Province, 261000, China.
Department of endocrinology, Weifang Sunshine Union Hospital, Weifang, Shandong Province, 261000, China.
BMC Musculoskelet Disord. 2024 Mar 25;25(1):235. doi: 10.1186/s12891-024-07349-6.
This study aimed to identify potential biomarkers for the diagnosis and treatment of osteoporosis (OP).
Data sets were downloaded from the Gene Expression Omnibus database, and differentially programmed cell death-related genes were screened. Functional analyses were performed to predict the biological processes associated with these genes. Least absolute shrinkage and selection operator (LASSO), support vector machine (SVM), and random forest (RF) machine learning algorithms were used to screen for characteristic genes, and receiver operating characteristics were used to evaluate the diagnosis of disease characteristic gene values. Gene set enrichment analysis (GSEA) and single-sample GSEA were conducted to analyze the correlation between characteristic genes and immune infiltrates. Cytoscape and the Drug Gene Interaction Database (DGIdb) were used to construct the mitochondrial RNA-mRNA-transcription factor network and explore small-molecule drugs. Reverse transcription real-time quantitative PCR (RT-qPCR) analysis was performed to evaluate the expression of biomarker genes in clinical samples.
In total, 25 differential cell death genes were identified. Among these, two genes were screened using the LASSO, SVM, and RF algorithms as characteristic genes, including BRSK2 and VPS35. In GSE56815, the area under the receiver operating characteristic curve of BRSK2 was 0.761 and that of VPS35 was 0.789. In addition, immune cell infiltration analysis showed that BRSK2 positively correlated with CD56dim natural killer cells and negatively correlated with central memory CD4 + T cells. Based on the data from DGIdb, hesperadin was associated with BRSK2, and melagatran was associated with VPS35. BRSK2 and VPS35 were expectably upregulated in OP group compared with controls (all p < 0.05).
BRSK2 and VPS35 may be important diagnostic biomarkers of OP.
本研究旨在鉴定骨质疏松症（OP）诊断和治疗的潜在生物标志物。
从基因表达综合数据库（GEO）下载数据集，筛选差异程序性细胞死亡相关基因。进行功能分析以预测与这些基因相关的生物学过程。最小绝对收缩和选择算子（LASSO）、支持向量机（SVM）和随机森林（RF）机器学习算法用于筛选特征基因，采用接收者操作特征（ROC）曲线评估疾病特征基因值的诊断效能。基因集富集分析（GSEA）和单样本 GSEA 用于分析特征基因与免疫浸润的相关性。利用 Cytoscape 和药物基因相互作用数据库（DGIdb）构建线粒体 RNA-mRNA-转录因子网络并探索小分子药物。采用反转录实时定量 PCR（RT-qPCR）分析评估临床样本中生物标志物基因的表达。
共鉴定出 25 个差异细胞死亡基因。其中，LASSO、SVM 和 RF 算法筛选出两个特征基因，包括 BRSK2 和 VPS35。在 GSE56815 中，BRSK2 的 ROC 曲线下面积为 0.761，VPS35 的 ROC 曲线下面积为 0.789。此外，免疫细胞浸润分析显示 BRSK2 与 CD56dim 自然杀伤细胞呈正相关，与中央记忆 CD4+T 细胞呈负相关。基于 DGIdb 数据，hesperadin 与 BRSK2 相关，melagatran 与 VPS35 相关。与对照组相比，OP 组中 BRSK2 和 VPS35 的表达均显著上调（均 P<0.05）。
BRSK2 和 VPS35 可能是 OP 的重要诊断生物标志物。
