Objective Bioinformatics method  was used to analyze gene expression microarrays of papillary thyroid cancer (PTC) and adjacent tissues. The key genes of PTC and their signal pathways were screened to understand their carcinogenic mechanisms.  Methods Data on PTC and paracancerous tissue samples were obtained from seven GSE series on two sequencing platforms in the Gene Expression Omnibus Database(GEO).Firstly, the differential genes of the two sequencing platform samples were screened by R language. Then biological function, signal pathway analysis and protein-protein interaction analysis were performed on differential genes by Metascape and STRING. Finally, the key gene were selcected by Cytoscape 3.5.1 software.  Results A total of 302 differential genes were obtained from the intersection of the two sequencing platform samples, of which 149 genes were up-regulated and 153 genes were down-regulated. Using the Cytoscape 3.5.1 software to screen out 15 key genes, 12 of them are involved in the extracellular matrix receptor interaction signal pathway. Survival analysis of 15 key genes was performed using the UALCAN database, and the changes in the expression levels of 4 genes were closely related to the survival time of patients.  Conclusion This study uses bioinformatics technology to analyze the data from seven PTC gene chips, making up for the inconsistency of small sample result  and improving the reliability and stability of the result . In addition, 15 key genes are screened out and found that the matrice extracellulaire receptor interactions pathway plays an important role in the development of thyroid cancer. The result  of this experiment provides guidance for the further study  of PTC molecular mechanism, diagnosis and screening of prognostic molecular markers.