Abstract
Lung adenocarcinoma (LUAD) is the most prevalent and lethal subtype of lung cancer worldwide and despite advances in diagnostic and therapeutic strategies, its prognosis remains poor. The present study aimed to identify key genes in LUAD through bioinformatics approaches. Transcriptomic data from the Gene Expression Omnibus and The Cancer Genome Atlas databases were analyzed using differential expression analysis, weighted gene co-expression network analysis, protein-protein interaction network construction and machine learning algorithms, and were validated using reverse transcription-quantitative PCR. Gene set enrichment analysis (GSEA) was performed to explore potential mechanisms associated with the involvement of key genes in LUAD, and single-cell transcriptomic data, collected from the Tumor Immune Single-cell Hub database, were used to validate cell-specific gene expression patterns. The results demonstrated that caveolin-1 (CAV1) and cadherin 5 (CDH5) are potential key genes in LUAD, both of which were significantly downregulated in tumor tissues compared with normal lung tissues. GSEA suggested that these genes are involved in the MAPK, Wnt and TGF-β signaling pathways, which are implicated in tumor progression. Furthermore, single-cell analysis revealed that CAV1 and CDH5 are predominantly expressed in endothelial cells, indicating a possible role in angiogenesis and tumor microenvironment regulation. In conclusion, CAV1 and CDH5 were systematically identified as potential tumor suppressor genes in LUAD, exhibiting robust diagnostic value confirmed by ROC analyses (GSE31210: CAV1 AUC=0.979; CDH5 AUC=0.969; GSE68465: 0.963 and 0.999; TCGA: 0.994 and 0.984). Therefore, CAV1 and CDH5 may serve as promising molecular targets for future therapeutic interventions, warranting further functional and clinical investigations.