Integrating single-cell RNA sequencing and spatial transcriptomics to reveal the Glycolysis-related gene GPRC5A as a potential biomarker for gastric cancer by machine learning.

BACKGROUND: Gastric cancer (GC) is a leading cause of cancer-related deaths due to late diagnosis. Altered glycolytic metabolism, notably the Warburg effect, plays a critical role in tumorigenesis, offering potential for early detection and targeted therapy.

METHODS: We combined single-cell omics and computational modeling to map and validate glycolytic biomarkers in GC. A novel interquartile range (IQR)-based approach was developed to assess glycolytic activity at the single-cell level, focusing on glycolysis-related genes (GRGs). The results were verified by in vitro and in vivo experiments.

RESULTS: We observed significant heterogeneity in glycolytic activity among GC cell subsets, with malignant cells exhibiting enhanced glycolysis. Nine hub genes-CD44, CLDN1, CLDN3, GPRC5A, IFNGR2, OLA1, PGK1, PLEK2, and TMEM147-showed strong diagnostic potential with high AUC values in ROC analysis. GPRC5A was selected for further validation, and its expression was significantly elevated in GC tissues and cell lines, confirming its role in glycolysis and GC progression.

CONCLUSIONS: This study proposes a glycolysis assessment strategy based on the interquartile range, which effectively addresses the limitations of previous scoring methods, including insufficient resolution and limited robustness. The proposed method enables more accurate quantification of intracellular glycolysis levels and facilitates fine-grained characterization of metabolic states. Furthermore, it identifies GPRC5A as a potential biomarker for the early diagnosis of gastric cancer. These findings enhance our understanding of metabolic reprogramming in gastric cancer and provide support for targeting the glycolytic pathway in therapeutic interventions.