OBJECTIVE: In both industrialized and developing nations worldwide, lung adenocarcinoma is one of the deadliest malignant tumors and the primary cause of cancer-related deaths. Its cellular heterogeneity is unclear to the fullest extent, although in recent years, its prevalence in younger individuals has increased. Therefore, it is urgent to deepen the understanding of lung adenocarcinoma and explore new therapeutic methods. METHODS: CytoTRACE, Monocle, SCENIC, and enrichment analysis were used to analyze the single cell RNA data, we characterized the biological characteristics of mast cells (MCs) in lung adenocarcinoma patient samples. CellChat was used to analyze and validate the interaction between MCs and tumor cells in lung adenocarcinoma. Prognostic models were used to evaluate and predict the development trend and outcome of a patient's disease, such as the survival time of cancer patients. The python package SCENIC was used to evaluate the enrichment of transcription factors and the activity of regulators in lung adenocarcinoma cell subgroups. CCK-8 assay could validate the activity of a specific cell subgroup sequenced in single cell sequencing to confirm the role of this cell subgroup in tumor proliferation. RESULTS: Our analysis identified seven major cell types, further grouping MCs within them and identifying four distinct subgroups, including MCs with high DUSP2 expression, which showed some tumor-related characteristics. In addition, we identified the key signaling receptor EGFR and validated it through in vitro knockdown experiments, demonstrating its role in promoting cancer. In addition, we established an independent prognostic indicator, the DUSP2+ MCs risk score, which showed an association between groups with high risk scores and poor outcomes. CONCLUSION: These findings shed light on the complex interactions in the lung adenocarcinoma tumor microenvironment and suggest that targeting specific MCs subgroups, particularly through the EGFR signaling pathway, may provide new therapeutic strategies.