Pancreatic adenocarcinoma (PAAD) is regarded as one of the most lethiferous cancers worldwide because treatment of pancreatic cancer remains challenging and mostly palliative. Little progress had been made to select certain reliable biomarkers as clinical prognosis. In this context, GSE28735 and GSE16515 were obtained from the Gene Expression Omnibus (GEO). GEO2R tool was used to recognize differentially expressed genes (DEGs). 351 DEGs were screened which included 230 up-regulated genes and 121 down-regulated genes. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to analyze the DEGs and associated signal pathways in the DAVID database. A protein-protein interaction (PPI) network was then constructed to screen 10 hub genes by STRING database and Cityscape software. Analyses of 10 hub genes were performed on GEPIA database and GSCA database, which revealed that MET was high expressed and significantly associated with survival of PAAD patients. Immunohistochemical staining showed that MET was higher expressed in PAAD tissues than adjacent tissues in 20 samples. The clinicopathological analysis revealed that high expression of MET was associated with the degree of differentiation, lymph node metastasis, vascular cancer thrombus and nerve invasion in PAAD tissues (P < .05). Furthermore, the Tumor Immune Estimation Resource (TIMER) database analyzed the correlation between the MET expression level and immune infiltration levels, which elucidated that MET expression was appreciably positively correlated with the infiltration levels of myeloid-derived suppressor cells (MDSCs). Here, these results strongly indicate MET is an unique prognostic biomarker. Its expression level is correlated with certain clinicopathological features and immune cell infiltration.