Identification of Three Prognosis-Related Differentially Expressed lncRNAs Driven by Copy Number Variation in Thyroid Cancer

Abstract

Thyroid cancer as the malignant tumor with the highest incidence in the endocrine system also shows a fast growth and development. In this work, we developed a new method to identify copy number variation– (CNV–) driven differentially expressed lncRNAs in thyroid cancer for predicting cancer prognosis. The data of RNA sequencing, CNV, methylation, mutation, and clinical details of thyroid cancer were obtained from the Cancer Genome Atlas database (TCGA). Molecular subtypes were clustered by iClusterPlus. Weighted gene co-expression network analysis (WGCNA) was employed to show co-expression modules. DEseq2 was conducted to identify protein coding genes (PCGs) and differentially expressed lncRNAs. CNV was detected using GISTIC 2.0. Three molecular subtypes were identified, and 68 differentially expressed lncRNAs (DElncRNAs) related to cancer were found among different molecular subtypes. CNV of FOXD2-AS1, FAM181A-AS1, and RNF157-AS1 was associated with overall survival and was involved in cancer-related pathways. These three DElncRNAs discovered based on CNV could serve as prognostic biomarkers to predict prognosis for thyroid cancer and new targets to explore molecular drugs.