In the current era of precision medicine, there is a general consensus that the anatomical site is an important factor in the management of colorectal cancer (CRC). To investigate the underlying molecular mechanisms between proximal and distal CRC and to identify the responsible genes, we analyzed the gene expression patterns of colorectal tumors from two microarray datasets, GSE39582 and GSE14333, on the NCBI Gene Expression Omnibus and the RNA‑seq data from TCGA. Weighted coexpression network analysis (WGCNA) was applied to construct a gene coexpression network. The red module in GSE39582 and the dark‑gray module from the TCGA dataset were found to be highly correlated with the anatomical site of CRC. A total of 12 hub genes were found in two datasets, 2 of which PLAG1 like zinc finger 2 (PLAGL2) and protein O‑fucosyltransferase 1 (POFUT1) were common and upregulated in tumor samples in CRC. The module with the highest correlation provided references that will help to characterize the difference between left‑sided and right‑sided CRC. The survival analysis of PLAGL2 and POFUT1 expression revealed differences between proximal and distal CRC. Gene set enrichment analysis based on those two genes provided similar results: GPI anchor biosynthesis and peroxisome and selenoamino acid metabolism. PLAGL2 and POFUT1, which have the highest correlation with tumor location, may serve as biomarkers and therapeutic targets for the precise diagnosis and treatment of CRC in the future.