Integrated Multi-omics Data Analysis and Validation Reveal the Crucial Role of Glycogen Metabolism in Gastric Cancer.

This study aimed to investigate glycogen metabolism in gastric cancer (GC) and develop a glycogen-based riskScore model for predicting GC prognosis. Patients' expression profiles for 33 tumor types were retrieved from TCGA. Four GC bulk and one single-cell sequencing datasets were obtained from GEO database. This study also enrolled a bladder urothelial carcinoma immunotherapeutic IMvigor210 cohort. The ssGSEA method was conducted to assess glycogen biosynthesis and degradation level. Consensus clustering analysis was conducted to identify different clusters. A glycogen riskScore signature was developed to evaluate prognostic value across different cohorts. Besides, experiments were conducted to further evaluate the role of glycogen metabolism related genes in GC. Both glycogen biosynthesis and degradation were significantly associated with worse overall survival and were also related with malignant phenotype in GC at both bulk and single-cell levels. Differential outcomes and immune functions were verified in the three identified clusters. The constructed glycogen riskScore model accurately classified GC patients with different outcomes, genomic and immune landscape, and performed well in predicting prognosis through external validation, immunotherapy and pan-cancer cohorts. Furthermore, the riskScore could predict response to chemotherapy and immunotherapy. Functional analyses revealed the signature's connection to pro-tumor and immunosuppression related pathways across pan-cancer. Additionally, glycogen metabolism related genes were found to regulate the malignant phenotypes of GC cells. This study revealed important roles of glycogen metabolism in promoting progression of GC and presented a glycogen riskScore model as a novel tool for predicting prognosis and treatment response.