Identification Of Anoikis-related Genes In Skin Cutaneous Melanoma Based On Bioinformatics Methods And Construction Of A Prognostic Model

Objective: Skin cutaneous melanoma(SKCM)is the most lethal type of skin cancer,and approximately 75% of skin cancer patients die from melanoma,which is highly aggressive.Melanoma has an unstable genome and is one of the most mutated cancers in somatic cells.Although it is possible to target mutated genes for treatment,more than half of patients are resistant to immunotherapy.Anoikis is programmed cell death,which is closely related to tumor metastasis.The purpose of this paper is to study the relationship between cutaneous melanoma and anoikis genes,find out the biomarkers related to anoikis of cutaneous melanoma,provide reference for clinical treatment of patients,and improve the prognosis of patients.Methods: Skin cutaneous melanoma data were obtained from TCGA and GEO databases.Anoikis-related genes(ARGs)were collected and identified.Univariate Cox analysis identified ARGs associated with SKCM.SKCM-ARG was typed by consensus clustering;GSEA analysis was used to clarify which functions and pathways the different subtypes were enriched in;ss GSEA was used to assess tumor infiltrating immune cell subsets between different subtypes.Lasso and multivariate Cox regression analysis were performed to obtain prognostic genes and build a prognostic model.Calculate the risk score,and then classify the patients into high-and low-risk categories based on that score.K-M analysis was performed to compare the survival difference between TCGA group and GEO group.ROC analysis evaluates the predictive performance of the model.The nomogram was drawn to predict patient survival,and DCA analysis was used to evaluate the accuracy of the nomogram.The differences in tumor microenvironment(TME)between the two groups were assessed.To predict the effectiveness of immune checkpoint inhibitors,chemotherapy drugs,and targeted drugs for both groups.TISCH provides cell-type annotations for model genes at the singlecell level.Results: Univariate Cox analysis yielded 133 ARGs associated with SKCM.Consensus clustering divided SKCM-ARG into two subtypes,and functional enrichment analysis showed that immune-related signaling pathways were significantly enriched,and the degree of immune cell infiltration in subtype A was higher than that in subtype B.A prognostic model based on 11 ARGs(DAPK2,ITGA6,CEBPB,PLK1,NOX4,MAPK10,PTK6,IFI27,NOTCH3,NDRG1,KRT14)was built,and patients were split into high-risk and low-risk according to the median value of the risk score risk group.Risk score was considered as a predictor in multivariate Cox regression analysis.According to the analysis of different subtypes of SKCM-ARG and the prognosis genes,five key genes were identified: NOX4,PTK6,IFI27,NOTCH3,KRT14.A nomogram with good accuracy has been established to provide more useful prognostic indicators for clinical practice.Risk score was negatively correlated with immune checkpoint inhibitors(anti-CLAT-4 and anti-PD-1).The low-risk group responded more sensitively to immunotherapy.Conclusion: The constructed prognostic model based on 11 genes can classify patients into high-risk and low-risk groups,with the low-risk group having a better survival rate than the high-risk group.SKCM-ARG was divided into subtypes A and B,and patients with subtype B were at higher risk;and 5 key genes were identified according to the prognostic genes and the two subtypes.A nomogram for detecting survival rate was drawn,which is expected to provide reference for clinical treatment.Patient risk groups can predict their sensitivity to different drug treatments,and patients in the low-risk group respond better to immune checkpoint inhibitors.