Mining novel biomarkers for prognosis of skin cutaneous melanoma with proteomic analysis

Abstract

Melanoma, a highly malignant tumor arising from melanocytes, poses a significant health threat with increasing incidence. This study leveraged proteomics, analyzing 352 samples from the Cancer Genome Atlas. We identified eight prognosis-related proteins (FOXO3A, CD171, CASPASE7CLEAVEDD198, Melanoma gp100, SRC, 1433ZETA, P21, and CABL) and constructed a prognostic model. The model accurately predicted patient outcomes, distinguishing high- and low-risk groups. Statistical analysis revealed no significant differences in clinical phenotypes between these groups. Principal Component Analysis validated model efficacy, and survival analysis indicated lower overall and progression-free survival in high-risk patients. Independent prognostic analysis and ROC curve analysis affirmed the model's reliability, with a higher predictive capacity than traditional clinical traits. Correlation analysis linked prognosis-related proteins with age, gender, and tumor stage. Our findings contribute valuable insights for diagnostic markers, drug targets, and understanding SKCM pathogenesis, advancing precision medicine. Limitations include the need for subgroup analysis and additional in vitro/in vivo validation.