A succinylation-related machine learning signature reveals prognostic and immune heterogeneity in lung adenocarcinoma

Succinylation is an emerging post-translational modification involved in tumor metabolism and progression. However, the role of succinylation-related genes (SRGs) in lung adenocarcinoma (LUAD) remains poorly understood. This study aimed to systematically investigate the prognostic and immunological significance of SRGs in LUAD. Transcriptomic and clinical data of LUAD were obtained from The Cancer Genome Atlas (TCGA). Weighted gene co-expression network analysis (WGCNA) and differential expression analysis were performed to identify LUAD-associated SRGs. Machine learning algorithms were applied to construct prognostic models, and Shapley Additive Explanations (SHAP) analysis was used to evaluate gene importance and model interpretability. Immune infiltration, immune checkpoint expression, pathway enrichment, and drug sensitivity analyses were conducted to characterize tumor heterogeneity. Immunohistochemistry was used to validate lactate dehydrogenase (LDHA) expression in LUAD tissues. A total of 156 succinylation-related differentially expressed genes were identified, among which 34 genes were significantly associated with overall survival. A machine learning-based prognostic signature effectively stratified LUAD patients into high- and low-risk groups with distinct survival outcomes. The high-risk group exhibited altered immune infiltration patterns, differential immune checkpoint expression, and distinct predicted drug sensitivities. SHAP analysis identified LDHA as the most important contributor to the prognostic model, and immunohistochemistry confirmed its high expression in LUAD tissues. We developed a SRGs prognostic model that reveals molecular and immune heterogeneity in LUAD. These findings provide potential biomarkers for prognosis prediction and therapeutic stratification in LUAD.