Integrated multi-omics and SHAP analysis reveal the core mechanisms of metabolic reprogramming in lung adenocarcinoma

Metabolic reprogramming is a pivotal hallmark of the malignant progression of lung adenocarcinoma (LUAD), yet its core regulatory genes and underlying molecular mechanisms remain largely elusive. In this study, we systematically deciphered the core regulatory network of LUAD metabolic reprogramming by integrating multi-omics analysis with the SHAP (SHapley Additive exPlanations) algorithm. Initially, LUAD transcriptomic and clinical data were acquired from the TCGA and GEO databases. Combined with 703 metabolic reprogramming-related genes retrieved from the Genecards database, 40 differentially expressed genes were identified via differential analysis, and key genes significantly impacting patient survival were subsequently isolated through prognostic analysis. Utilizing SHAP analysis to quantify the specific prognostic contributions of these genes, GPI, PFKP, and LDHB were recognized as the core regulatory genes. Single-cell sequencing analysis revealed that these three genes are highly expressed in the epithelial cells of LUAD tumor tissues and are closely associated with immune cell infiltration. In vitro cellular functional assays confirmed that silencing GPI, PFKP, or LDHB significantly restrained the proliferation and invasion capabilities of A549 cells, whilst regulating glucose metabolism and lactate production. Virtual knockout experiments further unraveled the downstream signaling pathway networks orchestrated by these three genes. Through combined multi-omics and SHAP analysis, this study elucidates, for the first time, the central roles of GPI, PFKP, and LDHB in the metabolic reprogramming of LUAD, providing a novel theoretical foundation for the development of diagnostic biomarkers and targeted therapeutics.