IntroductionLung adenocarcinoma, a disease with complex pathogenesis, high mortality and poor prognosis, is one of the subtypes of lung cancer. Hence, it is very crucial to find novel biomarkers as diagnostic and therapeutic targets for LUAD.MethodsGSE10072 was used for DEGs and WGCNA, and the intersection genes were subjected to enrichment analysis through Metascape and GSEA. Key genes were screened by three machine learning methods. Further, the reliability of key genes was identified by ROC, COX regression analysis and qRT-PCR. CIBERSORT and Spearman analysis were used for understanding the relationships of LUAD, immunity and key genes. In addition, ceRNA networks and potential drugs of key genes were constructed and predicted. ResultsAfter overlapping 631 DEGs and key module genes, 623 intersection genes were obtained. Subsequently, DUOX1, CD36, AGTR1, FHL5 and SSR4 were further selected using three machine learning methods. Reliability analysis demonstrated that AGTR1 possesses important predictive value for the occurrence and prognosis of LUAD. The enrichment analysis showed that AGTR1 was significantly enriched in the GPCR-related pathways. Immune infiltration analysis showed that the development of LUAD was related to the changes of immune cells such as M2 macrophages and neutrophils, which were regulated by AGTR1. Further, AGTR1 is also involved in regulating immune chemokines, checkpoints and immune regulatory factors such as PECAM1, ADARB1, SPP1 and ENO1, all of them playing important roles in immune cell regulation, tumor cell proliferation and migration. Further, the drug-gene interaction network screened out 13 potential drugs such as Benazepril, Valsartan, Eprosartan, and so on. DiscussionAGTR1 is a potential biomarker for the occurrence and progression of LUAD, closely related to tumor immunity, proliferation and migration. It can serve as a new target for the diagnosis and treatment of LUAD.