Background and Aims The incidence of thyroid cancer is increasing over years. Although its overall prognosis is favorable, some patients still die due to recurrence or metastasis. The purpose of this study was to screen the prognostic risk genes for thyroid carcinoma using bioinformatics approaches based on public databases.Methods The protein-coding gene RNA-seq data of thyroid cancer were downloaded from Cancer RNA-Seq Nexus (CRN) database and the differentially expressed protein-coding genes were screened. Then, enrichment analysis of the differentially expressed protein-coding genes was performed using the DAVID database. Protein-protein interaction networks among the differentially expressed protein-coding genes were constructed and analyzed using STRING and Cytoscape. The hub genes and their functional prediction were screened by the Cytohubba and ClueGO plugins, respectively. The expression level of hub genes was verified in thyroid cancer based on the UALCAN database, and survival analysis of hub genes was conducted in the GEPIA database to analyze whether their expression had an impact on the survival time of thyroid cancer.Results A total of 913 differentially expressed protein-coding genes were obtained after screening. These genes were mainly involved in regulation of small GTPase mediated signal transduction, Z disc, actin binding and drug metabolism-cytochrome P450. After construction of interaction networks, 10 hub genes were screened and they were TP53, ESR1, FOS, SYP, PPARG, ACTB, GRIA1, NRXN1, HDAC3 and KIT, of which TP53 had the highest score of 62. All of them were down-regulated in thyroid cancer tissue. Prediction results revealed that TP53, ESR1 and PPARG were probably involved in negative regulation of gene silencing, and TP53 and FOS were probably involved in the process of pri-miRNA transcription by RNA polymerase II. Results of verification in the UALCAN database showed that all except TP53, all other hub genes were down-regulated in thyroid cancer tissues (all P<0.05), which was consistent with the expression results in the CRN database. Results of survival analysis showed that high expression of KIT was significantly associated with disease-free survival of thyroid cancer patients (P=0.012), but had no significant effect on their overall survival (P=0.85).Conclusion The identified protein-coding gene KIT has a low expression in thyroid cancer tissue, and its high expression is closely associated with the disease-free survival of thyroid cancer, which is speculated to be a prognostic risk marker or therapeutic target for thyroid cancer.