Identification of a nine-miRNA signature for the prognosis of Uveal Melanoma


    loading  Checking for direct PDF access through Ovid

Abstract

The present study aims to construct a miRNA-based predictive signature of Uveal melanoma (UM) based on the database of the cancer genome atlas (TCGA). We obtained miRNA expression profiles and clinical information of 80 UM patients from TCGA, and randomly divided them into a training and a testing set. After data processing and forward screening, a total of 204 miRNAs with prognostic value were then examined by the Cox proportional hazard regression model in the training set. Receiver operating curve (ROC) analysis was applied to validate the accuracy of the signature. The biological relevance of putative miRNA target genes was also analyzed using the bioinformatics method. As a result, a linear prognostic model consisting of 9 miRNAs (miR-195, miR-224, miR-365a, miR-365b, miR-452, miR-4709, miR-7702, miR-513c, miR-873) was developed to divide UM patients into a high- and a low-risk group. Patients assigned to the high-risk group had significantly shorter overall survival than those in the low-risk group, which was further confirmed by the Area under curve (AUC) value of 0.858 at 5 year obtained from ROC. Gene Ontology (GO) analysis indicated that predicted target genes of these miRNAs are primarily associated with the modulation of protein expression and function, such as the activity of ubiquitin protein ligase and protein kinase. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis revealed that these genes were involved in multiple signaling pathways linked to carcinogenesis. The tumor specific 9-miRNA signature was also verified in the testing and entire set. In summary, based on UM data of TCGA, we identified and validated a 9-miRNA-based prognostic signature.HighlightsBased on TCGA, A 9-miRNA predictive signature of UM was established. The prognostic model works efficiently in sensitivity and specificity.We are able to score each patient according to these miRNAs and thus screen high-risk UM patients.

    loading  Loading Related Articles