Development and validation of autophagy-related-gene biomarker and nomogram for predicting the survival of cutaneous melanoma

Abstract

Autophagy plays a critical role in cutaneous melanoma, but the prognostic research of differentially expressed autophagy-related genes (DEARGs) in melanoma is lacking. Therefore, autophagy-related gene expression data of 630 melanoma patients were obtained from Gene Expression Omnibus (GEO) and The Cancer Genome Atlas database. The DEARGs were identified by “limma” package in R software. Best survival analysis and the least absolute shrinkage and selection operator method were performed to identify a robust 7-DEARG signature as well as construct nomogram associated with the clinical characteristics and validation in internal and external sets. This 7-DEARG signature could be regarded as an independent prognostic signature in clinical setting, and nomogram may supply a more simple and accurate prediction for the prognosis of melanoma. Moreover, 5 cancer hallmarks and 18 potential compounds are commonly enriched in high-risk expression phenotype. Thus, our finding provides new clinical evidences for the accurate diagnosis and identifies a potential prognostic autophagy-related marker for the treatment of melanoma.