BackgroundGemcitabine (GEM) is a second-line anticancer drug of choice for some colorectal cancer (CRC) patients, and GEM inability to be commonly available in the clinic due to the lack of clarity of the exact action targets.MethodsThe half maximal inhibitory concentration (IC50) of GEM treatment for 42 CRC cell lines were accessed from the Genomics of Drug sensitivity in Cancer (GDSC) database. High-throughput sequencing data of CRC patients were captured in The Cancer Genome Atlas (TCGA) and Weighted correlation network analysis (WGCNA) was conducted. Pearson correlations were derived for GEM potency-related genes. Differential analysis was conducted in the TCGA cohort to obtain CRC development-related genes (CDRGs), and univariate COX model analysis was performed on CDRGs overlapping with GEM potency-related genes to obtain CDRGs affecting CRC prognosis. Hub genes affecting GEM potency were identified by Spearman correlation.ResultsCALB2 and GPX3 were identified as potential targets for GEM treatment of CRC via prognostic analysis, which we also observed to be elevated with elevated clinical stage in CRC patients. The enhanced expression of CALB2 and GPX3 genes identified in the pathway analysis might inhibit the body metabolism as well as activate immune and inflammation related pathways. In addition, we found that CALB2 and GPX3 could also be considered as prognostic biomarkers in pan-cancer. Finally, we found that CALB2 and GPX3 were remarkably associated with the drug sensitivity of MG-132, Dasatinib, Shikonin, Midostaurin, MS-275, and Z-LNle-CHO, which were expected to be the drugs of choice for GEM combination.ConclusionCALB2 and GPX3 represent prognostic biomarkers for CRC and they might be potential action targets for GEM. Our study offered innovative ideas for GEM administration strategies.