| Purpose Thyroid cancer is the most prevalent malignancy in endocrine system,and differentiated thyroid cancer(DTC)accounts for over 90% of thyroid cancer.BRAF mutations occurs in around 45% of all DTC patients and normally implies poorer outcomes.However,prognosis varies greatly among them.Therefore,identifying novel biomarkers to predict progression and prognosis of BRAF-mutant DTC patients would be beneficial to risk stratification and therapeutic decisions for those patients.Methods The gene expression profiles and clinical information of DTC patients were obtained from The Cancer Genome Atlas database.Weighted gene co-expression network analysis was used to construct gene co-expression network regulated by BRAF mutations,and clinically significant gene modules in BRAF-mutant DTC was identified by moduletrait analysis.Biological functional pathways of gene modules were enriched by Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway analysis.Hub genes were selected from clinically significant gene modules and prognostic biomarkers for BRAF-mutant DTC patients were identified by disease-free survival analysis.Results 1047 differentially expressed genes which were regulated by BRAF mutations in DTC were identified.Three co-expression gene modules regulated by BRAF mutations in DTC were constructed,and two gene modules were associated with N grade of DTC.In addition,36 hub genes in these two modules were screened out,and 9 genes(SLC26A4,SOD3,KIT,GNAI1,FN1,ITGA2,IQGAP2,SORBS2,and ITGA3)were significantly correlated with prognosis of DTC.Further grouped diseasefree survival analyses were performed and 3 genes(FN1,ITGA3,and GNAI1)which exhibited reverse prognostic effect in BRAF-mutant and BRAF wild-type patients were found out.They specifically correlated with prognosis of DTC patients with BRAF mutations,but had no correlation with prognosis of DTC patients with wild-type BRAF.Finally,a combined survival analysis proved these 3 genes a robust gene panel exhibiting most significant correlation with prognosis of DTC patients with BRAF mutations.Conclusions Our study revealed two gene modules associated with cervical lymph nodes metastasis of DTC and 9 genes within two modules were identified as biomarkers for prognosis of DTC.Furthermore,2 genes(FN1 and ITGA3)and a gene panel(FN1,ITGA3,and GNAI1)were identified as biomarkers to predict prognosis specifically of BRAF-mutant DTC patients. |