Abstract: Objective To identify differentially expressed genes in colorectal cancer through bioinformatics, and to study the expression of candidate genes in colorectal cancer, as well as their pathological characteristics and prognostic significance. 
Methods Differentially expressed genes in colorectal cancer were identified by analyzing gene expression data sets from the GEO database, protein interaction networks was constructed, core proteins were found, gene set enrichment analysis was performed, and protein action pathways were clarified, then core protein expression levels between colorectal cancer tissues and normal tissues was compared by immunohistochemistry, finally the relationship between core genes and the prognosis of colorectal cancer was further clarified though survival analysis. 
Results A total of 89 differentially expressed genes were identified by screening criteria, and the core gene SERPINE1 was screened though constructing protein interaction network. Gene enrichment analysis showed that SERPINE1 expression was related to inflammation and angiogenesis. The high expression of SERPINE1 was significantly related to the invasion depth （P=0.012）, lymph node metastasis （P=0.025）, distant metastasis （P=0.040）, and TNM stage （P=0.016） of colorectal cancer. Survival analysis showed that high expression of SERPINE1 was associated with poor overall survival of colorectal cancer(P=0.002). 
Conclusion SERPINE1 may be an oncogene in the development of colorectal cancer, and high expression of SERPINE1 is a poor prognostic factor for colorectal cancer.

Key words: Colorectal Neoplasms, SERPINE1, immunohistochemistry, prognosis