5-hydroxymethylcytosine (5-hmC) as an epigenetics marker has significant impacts on cancer progression. Identification of preserved 5-hmC-related subnetworks in pan-cancer studies could lead to a better understanding of gastrointestinal (GI) cancers insights. Here, we conducted a network-based analysis on 5-hmC values of GI cancers, including colon, gastric, pancreatic cancers, and healthy donors. The co-5-hmC network was reconstructed using the weighted gene coexpression network (WGCNA) method. The hierarchical clustering method was implemented to detect pan-cancer-related modules/subnetworks. The preservation of modules was assessed using another dataset. Modules were functionally enriched, and biological pathways were visualized using the ConsensuspathDB. A 5-hmC predictive model was determined using the elastic network classifier to distinguish cancer patients and healthy individuals. To assess the efficiency of the model the recursive operating characteristics (AUC) curve was computed using the 5 cross-fold validation and an external dataset as well. Three pan-cancer-related subnetworks were detected preserved in another dataset. The main biological pathways were the cell cycle, apoptosis, and extracellular matrix (ECM) organization. The direct association between the cell cycle and ECM, the inverse association between apoptosis and ECM organization, and the inverse association between the cell cycle and ECM organization were detected for the 5-hmC marker in GI cancers. The AUC of 92% (0.73-1.00) was detected for the predictive model. In conclusion, the intricate association among biological pathways of ECM organization, Cell cycle, and apoptosis in GI cancers might be the consequence of epigenetics aberration; such findings could be beneficial in precision medicine using liquid biopsy in early stages.