Abstract
BACKGROUND: A growing number of evidence has revealed the vital role of autophagy in pathological processes of cancer, including gastric cancer (GC). However, many previous studies only focused on exploring single pathway or limited genes of interest in GC, which only reflected partial functions of autophagy. The present study aimed to construct an autophagy-related risk signature for GC. METHODS: Differentially expressed autophagy-related genes (ARGs) in GC and non-tumor samples were screened through The Cancer Genome Atlas (TCGA) database, followed by bioinformatics analysis using the Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) platforms. Prognosis-related ARGs were generated by univariate and multivariate Cox regression test. RESULTS: A total of seven prognosis-related ARGs (HSPB8, NRG2, GABARAPL1, TMEM74, DLC1, MAP1LC3C and NRG3) were determined to establish a prognostic index (PI) model, which was demonstrated to be an independent prognostic indicator for patients with GC. More importantly, it was successfully validated in an external cohort of patients from the GSE15460 dataset, indicating the useful reproducibility of this signature. In addition, the PI model was associated with immune cell infiltration estimates in GC. CONCLUSIONS: Taken together, the present study suggested that the seven ARGs-related signature could serve as an independent prognostic indicator for patients with GC.