Uterine corpus endometrial carcinoma (UCEC) is one of the most common gynecologic malignancies, but only a few biomarkers have been proven to be effective in clinical practice. Previous studies have demonstrated the important roles of non-coding RNAs (ncRNAs) in diagnosis, prognosis, and therapy selection in UCEC and suggested the significance of integrating molecules at different levels for interpreting the underlying molecular mechanism. In this study, we collected transcriptome data, including long non-coding RNAs (lncRNAs), microRNAs (miRNAs), and messenger RNAs (mRNAs), of 570 samples, which were comprised of 537 UCEC samples and 33 normal samples. First, differentially expressed lncRNAs, miRNAs, and mRNAs, which distinguished invasive carcinoma samples from normal samples, were identified, and further analysis showed that cancer- and metabolism-related functions were enriched by these RNAs. Next, an integrated, dysregulated, and scale-free biological network consisting of differentially expressed lncRNAs, miRNAs, and mRNAs was constructed. Protein-coding and ncRNA genes in this network showed potential immune and metabolic functions. A further analysis revealed two clinic-related modules that showed a close correlation with metabolic and immune functions. RNAs in the two modules were functionally validated to be associated with UCEC. The findings of this study demonstrate an important clinical application for improving outcome prediction for UCEC.
Keywords: TCGA; dysregulated network; endometrial carcinoma; immunity; integrative analysis; lncRNA; metabolism; miRNA.
Copyright © 2021 Liu and Qiu.