Abstract
Background: Acute myeloid leukemia (AML) is a type of blood cancer with diverse genetic pathogenesis. The identification of potential molecular biomarkers could improve the AML diagnosis and targeted therapy. Recently, competing endogenous RNAs (ceRNAs) became an active area in cancer research to determine molecular mechanisms underlying tumor development. Objectives: The aim of the present study was to investigate the key molecular biomarkers that are closely related to AML through bioinformatics analysis. Methods: In this research, the RNA-seq data of 151 AML patients and 151 corresponding health samples were retrieved from The Cancer Genome Atlas (TCGA) database and GTEx Portal (genotype-tissue expression), respectively. After that we screened the differentially expressed long non-coding RNAs (lncRNAs), microRNAs (miRNAs), and mRNAs by “limma” package in R to construct the co-expression network ceRNA (miRNA-lncRNA-mRNA) with weighted gene co-expression network analysis (WGCNA) package in R and Cytoscape (version 3.7.2) software, respectively. Then the relevant modules were identified and functional enrichment analysis was used to uncover the modules that are biologically related to AML cancer. Results: Based on our bioinformatics studies, we constructed a significant module which contain 333 genes. Among them, five up-regulated miRNAs with the highest GS (gene significant), include hsa-miR-374B, hsa-miR-553, hsa-miR-3679, hsa-miR-548L, hsa-miR-597 and five down-regulated miRNAs with the lowest GS including hsa-miR-3934, hsa-miR-4746, hsa-miR-466, hsa-miR-6722, hsa-miR-4490. For protein-coding genes (PCGs), the top-five up-regulated PCGs with the highest GS were AMIGO3, H2AC17, SPACA5, GPR21, and OR13C3. The top-five down-regulated PCGs with the lowest GS were XBP1, TOMM6, TREX1, TNFRSF6B, and BOLA2. Among the lncRNAs, the five up-regulated lncRNAs with the highest level of GS were GREP1, ZBTB20-AS2, LINC01596, LINC00345, and RMRP. The five down-regulated lncRNAs with the lowest GS were LINC01609, LINC01707, LINC02270, LINC02309, and LINC02243. These lncRNAs can serve as potential biomarkers to distinguish between AML and normal samples. Conclusions: Finally, considering the role of ceRNA network in various biological processes, examining the role of ceRNAs in AML may provide a deeper insight into molecular mechanisms underlying the pathogenesis of cancer. This understanding may propose potential biomarkers for diagnosis and therapeutic interventions.