Bioinformatics Analysis Of Iron Death-related Gene Expression Differences And Clinical Characteristics In Lung Adenocarcinoma

Objective: To link GEO lung adenocarcinoma gene expression profile analysis with iron death through biological information analysis,screen molecular markers related to the diagnosis of lung adenocarcinoma,and further explore the correlation of diagnostic markers in survival analysis and clinical characteristics of lung adenocarcinoma.Methods: 1.Differential analysis of genes was conducted by GEO database using LIMMA,and the differential genes related to lung adenocarcinoma were obtained.Ferr Db database was used to download iron death related genes,and the intersection of the two genes was obtained to obtain differential iron death genes.The mechanism of action was predicted by GO and KEGG functional enrichment analysis,and the PPI protein interaction network was constructed.2.Lasso and SVM-RFE regression analysis were used to screen model genes related to lung adenocarcinoma,draw Receiver Operating Characteristic Curve(ROC)and calculate the Area Under Curve of each indicator.AUC)and validate in the external validation set;3.Subsequently,model gene expression was verified and survival analysis was performed in dataset GSE30219,and correlation analysis of clinical features was performed based on clinical data of patients in TCGA database.Results: 1.GSE30219 differential analysis screened out 1490 differential genes,including 493 up-regulated genes and 997 down-regulated genes.32 differential iron death genes were obtained by intersection of the abovementioned differential genes with 258 iron death genes.GO and KEGG enrichment analysis showed that the role of these genes may be related to arachidonic acid metabolism and oxidative stress.2.Lasso+SVM-RFE analysis screened out three characteristic genes: CDO1,IL33 and NOX4.In the evaluation of its diagnostic effect in the external validation set(GSE31210),its AUC value was 0.967,indicating that the model performed well.3.In the verification of the expression level of model genes,CDO1 and IL33 were highly expressed in normal samples,and low expression in lung adenocarcinoma samples;NOX4 was low expression in normal samples,and high expression in lung adenocarcinoma samples.Survival analysis showed that CDO1,IL33 and NOX4 were significant between high and low gene expression groups.The expression level of CDO1 gene was significantly different in STAGEI-II,STAGEI-III and STAGEI-IV stages,T1-T2,T1-T3 and T1-T4 stages,and there were significant differences between smoking pack number <30 and >=30groups.There was no significant difference in age <60 group and >60group,different sex group,different N and M stage.The expression level of IL33 gene was significantly different in STAGEI-II,T1-T2 and N1-N2 stages,and there were significant differences between age <60 and >60groups,and between smoking pack number <30 and >=30 groups.There was no significant difference in different sex groups and M stage.NOX4 gene and lung adenocarcinoma clinical features are not significantly different.Conclusions: this study is based on using R language to establish and verify the GEO database based on iron death related gene diagnostic markers,and further explores its in lung adenocarcinoma of survival analysis and the correlation of clinical features,the results showed the CDO1,IL33,NOX4 the diagnosis of lung adenocarcinoma has a certain value,is screening potential molecular markers of lung adenocarcinoma.