Analysis of Autophagy-Related Signatures Identified Two Distinct Subtypes for Evaluating the Tumor Immune Microenvironment and Predicting Prognosis in Ovarian Cancer

Abstract

Ovarian cancer (OC) is one of the most lethal gynecologic malignant tumors. The interaction between autophagy and the tumor immune microenvironment has clinical importance. Hence, it is necessary to explore reliable biomarkers associated with autophagy-related genes (ARGs) for risk stratification in OC. Here, we obtained ARGs from the MSigDB database and downloaded the expression profile of OC from TCGA database. The k-means unsupervised clustering method was used for clustering, and two subclasses of OC (cluster A and cluster B) were identified. SsGSEA method was used to quantify the levels of infiltration of 24 subtypes of immune cells. Metascape and GSEA were performed to reveal the differential gene enrichment in signaling pathways and cellular processes of the subtypes. We found that patients in cluster A were significantly associated with higher immune infiltration and immune-associated signaling pathways. Then, we established a risk model by LASSO Cox regression. ROC analysis and Kaplan-Meier analysis were applied for evaluating the efficiency of the risk signature, patients with low-risk got better outcomes than those with high-risk in overall survival. Finally, ULK2 and GABARAPL1 expression was further validated in clinical samples. In conclusion, Our study constructed an autophagy-related prognostic indicator, and identified two promising targets in OC.