Molecular Network-based Identification Of Thyroid Carcinoma-related CeRNAs And Drugs

Thyroid cancer is the most usual malignancy of endocrine organs,accounting for about 1% of systemic malignant tumors.It has become the fifth most common cancer in women.Meanwhile,long non-coding RNAs(lncRNAs),which may act as competing endogenous RNAs(ceRNAs),interact with coding RNAs by competing same miRNA response elements(MREs),and play an important role in the regulation of gene expression in a variety of tumors.Therefore,it has become an important research direction to construct the network structure of three RNAs to study the mechanism of cancer.In this paper,based on the gene expression and clinical data in TCGA database,we proposed a network construction pipeline,and analysis of the drug prediction on the differentially expressed genes.The main work is represented as follows:1.We performed a thorough literature review on current research status of thyroid cancer,progress of lncRNAs,and mechanisms of ceRNAs,which lays a powerful foundation for followup research.2.We have developed a novel pipeline called Molecular Network-based Identification of ceRNA(MNIceRNA)to identify ceRNAs in thyroid carcinoma.MNIceRNA first performs differential RNA analyses using edgeR,and then constructs lncRNA-miRNA-mRNA networks from miRcode database and WGCNA,based on which to identify key drivers of differentially expressed RNAs between normal and tumor samples.Finally,MNIceRNA focuses on thyroid carcinoma associated key driver genes(KDGs)and constructs a ceRNA network according to the lnCeDB.As a result,598 lncRNAs,1025 mRNAs,and 90 miRNAs were inferred to be differentially expressed genes and then mapped to the network,we obtained 8 key driver miRNAs,among which hsa-mir-375 was inferred to be significant for patients’ survival with 11 of 34 associated ceRNAs,among which RUNX2,DUSP6 and SEMA3 D are known oncogenes regulating cellular proliferation and differentiation in thyroid cancer.3.We proposed an identification method of drugs or genes perturbations.We constructed two WGCNA networks from normal and diseased samples,respectively,and obtained 33787 significant differential co-expressions(P-value≤1.0E-20).And the differentially expressed RNAs are then used to query drug or gene perturbation expression signatures to identify interventions that were able to “reverse” the expression changes seen in the tumor sub-network.Finally,for the top ranked drugs or genes,we perform literature based evaluation for their roles or conduct inference on the potential underlying mechanisms,and obtained 1274 unique prioritized genes and drug targeted genes(P<0.001),based on which to predict thyroid cancer drugs for further experimental validation.