Crohn's disease (CD) is a segmental chronic inflammatory bowel disease, which seriously affects the patient's quality of life. The etiology of CD is not yet clear, and there is still a lack of effective treatments. Therefore, in this study, we focus on developing a useful model for early diagnosis and targeted therapy of CD. The expression datasets of CD were collected to filter differentially expressed genes (DEGs) by overlapping "limma" package and "WGCNA" package. Then, functional enrichment analysis and protein-protein interaction (PPI) network analyses were performed. Hub genes were screened with "cytoHubba" plug-in and filtered with LASSO and stepwise regression analyses. The logistic regression model and nomogram were established based on the selected hub genes. The 45 DEGs were identified and the top 30 hub genes were chosen out for further study. Finally, 11 genes were selected to construct the logistic regression model and nomogram. The receiver operating characteristic (ROC) curve shows that the area under the curve (AUC) value was 0.960 in the training dataset and 0.760 in the validation dataset. A 11-gene diagnostic model was constructed with IL1B, CXCL10, CXCL2, LCN2, MMP12, CXCL9, NOS2, GBP5, FPR1, GBP4 and WARS, which may become potential biomarkers for early diagnosis and targeted therapy of CD.