Data mining of immune-related prognostic genes in metastatic melanoma microenvironment

Ubiquitination-related biomarkers in metastatic melanoma patients and their roles in tumor microenvironment

Background

Skin cutaneous melanoma (SKCM) is the deadliest type of cutaneous malignancy. Ubiquitination is a process of protein sorting and degradation that exhibits multiple functions in the progression of various tumors. This study aimed to characterize a set of genes for ubiquitination in SKCM.

Methods

The expression patterns of ubiquitin-associated genes (URGs) and the corresponding clinical information in SKCM tissues were comprehensively analyzed based on The Cancer Genome Atlas (TCGA) database. We performed univariate and multivariate Cox proportional regression models to characterize the risk scores and identify four critical genes related to prognostic ubiquitination (HCLS1, CORO1A, NCF1 and CCRL2), which were used to construct the prognostic signatures. We also studied the effects of HCLS1, CORO1A and CCRL2 on tumor metastasis-related indicators at the cellular level through in vitro experiments.

Results

SKCM patients in the low-risk group showing a longer survival than those in the high-risk group. Characteristic risk scores correlated with several clinicopathological variables and reflected the infiltration of multiple immune cells. In addition, the knockdown of CLS1, CORO1A and CCRL2 affected cellular malignant biological behavior through the EMT signaling pathway.

Conclusion

This study provides a novel and prospective strategy to improve the clinical survival of SKCM patients.

Revisiting miRNA Association with Melanoma Recurrence and Metastasis from a Machine Learning Point of View

The diagnostic and prognostic value of miRNAs in cutaneous melanoma (CM) has been broadly studied and supported by advanced bioinformatics tools. From early studies using miRNA arrays with several limitations, to the recent NGS-derived miRNA expression profiles, an accurate diagnostic panel of a comprehensive pre-specified set of miRNAs that could aid timely identification of specific cancer stages is still elusive, mainly because of the heterogeneity of the approaches and the samples. Herein, we summarize the existing studies that report several miRNAs as important diagnostic and prognostic biomarkers in CM. Using publicly available NGS data, we analyzed the correlation of specific miRNA expression profiles with the expression signatures of known gene targets. Combining network analytics with machine learning, we developed specific non-linear classification models that could successfully predict CM recurrence and metastasis, based on two newly identified miRNA signatures. Subsequent unbiased analyses and independent test sets (i.e., a dataset not used for training, as a validation cohort) using our prediction models resulted in 73.85% and 82.09% accuracy in predicting CM recurrence and metastasis, respectively. Overall, our approach combines detailed analysis of miRNA profiles with heuristic optimization and machine learning, which facilitates dimensionality reduction and optimization of the prediction models. Our approach provides an improved prediction strategy that could serve as an auxiliary tool towards precision treatment.

Identification and validation of an epithelial mesenchymal transition-related gene pairs signature for prediction of overall survival in patients with skin cutaneous melanoma

BACKGROUND

We aimed to construct a novel epithelial-mesenchymal transition (EMT)-related gene pairs (ERGPs) signature to predict overall survival (OS) in skin cutaneous melanoma (CM) patients.

METHODS

Expression data of the relevant genes, corresponding clinicopathological parameters, and follow-up data were obtained from The Cancer Genome Atlas database. Univariate Cox regression analysis was utilized to identify ERGPs significantly associated with OS, and LASSO analysis was used to identify the genes used for the construction of the ERGPs signature. The optimal cutoff value determined by the receiver operating characteristic curve was used to classify patients into high-risk and low-risk groups. Survival curves were generated using the Kaplan-Meier method, and differences between the two groups were estimated using the log-rank test. The independent external datasets GSE65904 and GSE19234 were used to verify the performance of the ERGPs signature using the area under the curve (AUC) values. In addition, we also integrated clinicopathological parameters and risk scores to develop a nomogram that can individually predict the prognosis of patients with CM.

RESULTS

A total of 104 ERGPs related to OS were obtained, of which 21 ERGPs were selected for the construction of the signature. All CM patients were stratified into high-and low-risk groups based on an optimal risk score cutoff value of 0.281. According to the Kaplan-Meier analysis, the mortality rate in the low-risk group was lower than that in the high-risk group in the TCGA cohort (P < 0.001), GSE65904 cohort (P = 0.006), and GSE19234 cohort (P = 0.002). Multivariate Cox regression analysis indicated that our ERGP signature was an independent risk factor for OS in CM patients in the three cohorts (for TCGA: HR, 2.560; 95% CI [1.907-3.436]; P < 0.001; for GSE65904: HR = 2.235, 95% CI [1.492-3.347], P < 0.001; for GSE19234: HR = 2.458, 95% CI [1.065-5.669], P = 0.035). The AUC value for predicting the 5-year survival rate of patients with CM of our developed model was higher than that of two previously established prognostic signatures. Both the calibration curve and the C-index (0.752, 95% CI [0.678-0.826]) indicated that the developed nomogram was highly accurate. Most importantly, the decision curve analysis results showed that the nomogram had a higher net benefit than that of the American Joint Committee on Cancer stage system.

CONCLUSION

Our study established an ERGPs signature that could be potentially used in a clinical setting as a genetic biomarker for risk stratification of CM patients. In addition, the ERGPs signature could also predict which CM patients will benefit from PD-1 and PD-L1 inhibitors.