Multi-omics analysis reveals lysosome-associated molecular subtype characterization and prognostic modeling system in lung adenocarcinoma

<b>Background:</b> Lung adenocarcinoma (LUAD) poses a significant challenge in current treatments due to its high recurrence and metastasis rates. Despite preliminary evidence suggesting the role of lysosomes in LUAD, it remains unclear whether lysosome-related functions can be effectively used for risk stratification of LUAD patients and involved lysosome-related functional targets are still needed to be explored. <b>Method:</b> An integrated analysis of TCGA and GEO databases was conducted to explore the potential role of lysosome-related genes (LRGs) in LUAD. Unsupervised consensus clustering analysis was utilized to explore the LRG molecular subtypes in LUAD. ESTIMATE and ssGSEA algorithms were performed to evaluate the immune infiltration characterization of LUAD samples. LASSO-univariate and multivariate Cox analysis were used to construct the LRG score model. Single-cell sequencing analysis was performed to reveal the distribution characteristics in different cell subpopulations of selected LRGs. <i>In vitro</i> experiments including western blotting, PCR, colony formation assays, and Transwell assays were used to verify the expression and biological functions of the selected target in LUAD. <b>Results:</b> Through multi-omics integration analysis algorithms, we successfully developed a prognostic risk stratification system based on LRG scoring in LUAD and constructed a nomogram diagnostic model. Various bioinformatics analyses indicated the potential clinical value of the LRG scoring system. Single-cell sequencing analysis further revealed the composition of cell subpopulations and the expression characteristics of prognostic signatures. SLC2A1, one of the selected targets, was validated through <i>in vitro</i> experiments to regulate the proliferation and migration of LUAD cells, thereby confirming the reliability of the bioinformatics results. <b>Conclusion:</b> Our results demonstrate that effective risk stratification of LUAD patients can be achieved through LRGs by multi-omics analysis integration. Furthermore, we validated key prognostic targets <i>in vitro</i>, providing new ideas for future clinical treatment.