Identification of the Signature Genes and Network of Reactive Oxygen Species Related Genes and DNA Repair Genes in Lung Adenocarcinoma.

Reactive Oxygen Species (ROS) are present in excess amounts in patients with tumors, and these ROS can kill and destroy tumor cells. Therefore, tumor cells upregulate ROS-related genes to protect them and reduce their destructing effects. Cancer cells already damaged by ROS can be repaired by expressing DNA repair genes consequently promoting their proliferation. The present study aimed to identify the signature genes of and regulating network of ROS-related genes and DNA repair genes in lung adenocarcinoma (LUAD) using transcriptomic data of public databases. The LUAD transcriptome data in the TCGA database and gene expressions from Gene Expression Omnibus (GEO) were analyzed and samples were clustered into 5 ROS-related categories and 6 DNA repair categories. Survival analysis revealed a significant difference in patient survival between the two classification methods. In addition, the samples corresponding to the two categories overlap, thus, the gene expression profile of the same sample with different categories and survival prognosis was further explored, and the connection between ROS-related and DNA repair genes was investigated. The interactive sample recombination classification was used, revealing that the patient's prognosis was worse when the ROS-related and DNA repair genes were expressed at the same time. The further research on the potential regulatory network of the two categories of genes and the correlation analysis revealed that ROS-related genes and DNA repair genes have a mutual regulatory relationship. The ROS-related genes namely NQO1, TXNRD1, and PRDX4 could establish links with other DNA repair genes through the DNA repair gene NEIL3, thereby balancing the level of ROS. Therefore, targeting ROS-related genes and DNA repair genes might be a promising strategy in the treatment of LUAD. Finally, a survival prognostic model of ROS-related genes and DNA repair genes was established (TERT, PRKDC, PTTG1, SMUG1, TXNRD1, CAT, H2AFX, and PFKP). The risk score obtained from our survival prognostic model could be used as an independent prognostic factor in LUAD patients.