Construction And Validation Of A Prognostic Risk Model For Cutaneous Melanoma Based On Protein Expression Profiling

Objective:Skin Cutaneous melanoma(SKCM)is one of the most malignant skin cancers.Although the current treatment methods have improved compared with the previous ones,the prognosis is still poor,especially for metastatic cutaneous malignant melanoma.Combining advanced genomic analysis with proteomic characteristics to construct protein prediction models will help to screen effective biomarkers and may provide new therapeutic references for clinical treatment.Methods:This study obtained proteome data from the Cancer Proteome Atlas(TCPA)models,with GSEA enrichment analysis studying model protein significantly enriched pathways,and using the CIBERSORT tool to investigate the potential correlation between immune cells and gene features.Drug sensitivity analysis explores chemicals for the treatment of SKCM.Results:The model consists of nine proteins(Connexin-43,4EBP1＿p T70,ARAF,PCADHERIN,IDO,1433 ZETA,D-a-Tubulin,Glucocorticoid-Receptor and YAP).Protein prognostic models can be considered as independent factors based on risk curves,survival curves,ROC curves,and independent prognostic analysis to accurately predict the survival time of SKCM patients.The CIBERSOR algorithm was used to study the content of the model proteins in the 22 immune cells.Furthermore,we verified the expression of nine differential proteins by immunohistochemistry.Conclusions:This study suggests that prognostic-related proteins may serve as novel biomarkers for SKCM diagnosis,and that risk models can be used to predict the prognosis of SKCM patients.