Background: Lung cancer is one of the malignancies exhibiting the fastest increase in morbidity and mortality, but the cause is not clearly understood. The goal of this investigation was to screen and identify relevant biomarkers of lung cancer.

Methods: Publicly available lung cancer data sets, including GSE40275 and GSE134381, were obtained from the GEO database. The repeatability test for data was done by principal component analysis (PCA), and a GEO2R was performed to screen differentially expressed genes (DEGs), which were all subjected to enrichment analysis. Protein-protein interactions (PPIs), and the significant module and hub genes were identified via Cytoscape. Expression and correlation analysis of hub genes was done, and an overall survival analysis of lung cancer was performed. A receiver operating characteristic (ROC) curve analysis was performed to test the sensitivity and specificity of the identified hub genes for diagnosing lung cancer.

Results: The repeatability of the two datasets was good and 115 DEGs and 10 hub genes were identified. Functional analysis revealed that these DEGs were associated with cell adhesion, the extracellular matrix, and calcium ion binding. The DEGs were mainly involved with ECM-receptor interaction, ABC transporters, cell-adhesion molecules, and the p53 signaling pathway. Ten genes including COL1A2, POSTN, DSG2, CDKN2A, COL1A1, KRT19, SLC2A1, SERPINB5, DSC3, and SPP1 were identified as hub genes through module analysis in the PPI network. Lung cancer patients with high expression of COL1A2, POSTN, DSG2, CDKN2A, COL1A1, SLC2A1, SERPINB5, and SPP1 had poorer overall survival times than those with low expression (p <0.05). The CTD database showed that 10 hub genes were closely related to lung cancer. Expression of POSTN, DSG2, CDKN2A, COL1A1, SLC2A1, SERPINB5, and SPP1 was also associated with a diagnosis of lung cancer (p<0.05). ROC analysis showed that SPP1 (AUC = 0.940, p = 0.000*, 95%CI = 0.930-0.973, ODT = 7.004), SLC2A1 (AUC = 0.889, p = 0.000*, 95%CI = 0.791-0.865, ODT = 7.123), CDKN2A (AUC = 0.730, p = 0.000*, 95%CI = 0.465-1.000, ODT = 6.071) were suitable biomarkers.

Conclusion: Microarray technology represents an effective method for exploring genetic targets and molecular mechanisms of lung cancer. In addition, the identification of hub genes of lung cancer provides novel research insights for the diagnosis and treatment of lung cancer.

背景：肺癌是发病率和死亡率增加最快的恶性肿瘤之一，但其原因尚不清楚。本次调查的目的是筛选和鉴定肺癌的相关生物标志物。

方法：公开可用的肺癌数据集，包括 GSE40275 和 GSE134381，来自 GEO 数据库。数据的重复性测试采用主成分分析（PCA），GEO2R筛选差异表达基因（DEGs），均进行富集分析。蛋白质-蛋白质相互作用 (PPI) 以及重要的模块和中心基因通过 Cytoscape 进行鉴定。对hub基因进行表达及相关分析，对肺癌进行总体生存分析。进行受试者工作特征 (ROC) 曲线分析以测试已鉴定的枢纽基因诊断肺癌的敏感性和特异性。

结果：两个数据集的重复性良好，鉴定出115个DEGs和10个hub基因。功能分析表明，这些 DEGs 与细胞粘附、细胞外基质和钙离子结合有关。DEG 主要与 ECM 受体相互作用、ABC 转运蛋白、细胞粘附分子和 p53 信号通路有关。通过PPI网络中的模块分析，将COL1A2、POSTN、DSG2、CDKN2A、COL1A1、KRT19、SLC2A1、SERPINB5、DSC3和SPP1等10个基因确定为枢纽基因。COL1A2、POSTN、DSG2、CDKN2A、COL1A1、SLC2A1、SERPINB5和SPP1高表达的肺癌患者的总生存时间比低表达者差（p <0.05）。CTD数据库显示10个hub基因与肺癌密切相关。POSTN、DSG2、CDKN2A、COL1A1、SLC2A1、SERPINB5 和 SPP1 也与肺癌的诊断相关（p<0.05）。ROC 分析显示 SPP1 (AUC = 0.940, p = 0.000*, 95%CI = 0.930-0.973, ODT = 7.004), SLC2A1 (AUC = 0.889, p = 0.000*, 95%CI = 0.835, ODT = 0.8371. )、CDKN2A (AUC = 0.730, p = 0.000*, 95%CI = 0.465-1.000, ODT = 6.071) 是合适的生物标志物。

结论：微阵列技术是探索肺癌遗传靶点和分子机制的有效方法。此外，肺癌枢纽基因的鉴定为肺癌的诊治提供了新的研究思路。