Based on comprehensive analysis of ScRNA-seq and bulk RNA-seq, identification of palmitoylation-related genes as biomarkers to evaluate prognosis and immune landscape in stomach adenocarcinoma
摘要
Stomach adenocarcinoma (STAD) poses a significant global health burden, with notable geographic variations in its incidence and complex underlying causes. Palmitoylation plays a crucial role in cancer development, yet its prognostic relevance in STAD remains poorly understood. Multi-omics data integration was performed using TCGA-STAD (N=407) and GEO-GSE26901 (N=110) cohorts with single-cell data (GSE167297; N=14 samples). Palmitoylation-related genes (PRGs=3,599) from GeneCards underwent differential expression screening and Cox regression to identify prognostic DEPRGs. A LASSO-multivariate Cox model constructed a 5-gene signature for risk stratification, validated externally. Comprehensive analyses included molecular subtyping (ConsensusClusterPlus), immune infiltration (CIBERSORT/ssGSEA), immunotherapy response (TIDE/IPS), drug sensitivity (pRRophetic/CellMiner), pathway enrichment (GSEA/clusterProfiler), and single-cell ecosystem characterization. Based on multi-omics analysis of STAD, we identified 5 signature gene (GAMT, RDH12, RNASE1, SERPINE1, and SLC52A3) to construct a robust prognostic model. High-risk group exhibited enrichment in pro-metastatic pathways (CAMs, ECM-receptor interaction) and an immunosuppressive microenvironment (increased M2 macrophages, decreased CD8⁺ T cells), correlating with poor survival, higher TIDE scores, and immunotherapy resistance. Consensus clustering revealed two molecular subtypes with divergent survival outcomes. Single-cell analysis revealed SERPINE1⁺ endothelial cells as a high-risk subpopulation driving angiogenesis via MK pathway interactions with epithelial/stromal cells. Based on multi-omics integration, this study establishes PRGs signature as a robust prognostic biomarker for STAD, enabling risk stratification, molecular subtyping, and prediction of immunotherapy response while elucidating PRGs-mediated remodeling of the tumor-immune microenvironment.