Background <p>The role of mast cells (MCs) in colorectal cancer (CRC) has been a subject of debate, and their impact on liver metastasis remains largely unknown. This study aimed to determine the role of MCs and enriched genes in CRC with liver metastasis through a multi-omics analysis approach.</p> Methods <p>The proportion of MCs in CRC bulk sequencing datasets was characterized using the CIBERSORT algorithm, and the number of MCs in scRNA-seq datasets was analyzed. Associations between the proportion of MCs and clinical features of CRC were determined. MC- enriched genes were identified and used to construct a predictive model for liver metastasis and a prognostic model for CRC survival. The expression of MC- enriched genes was validated using external datasets and clinical samples via immunohistochemistry (IHC).</p> Results <p>Four bulk-seq datasets with 223 samples confirmed that the proportion of activated MCs was lower in liver metastasis tissues compared to primary CRC tissues. The scRNA-seq dataset revealed a smaller number of MCs in liver metastasis tissues versus primary CRC tissues, and the expression of activated MC genes was higher in primary CRC tissues compared to non-metastatic tissues. No significant association was found between the proportion of activated MCs and the TNM stage or tumor stage of CRC. Three genes (APOC1, SPP1, CLU) were identified by overlapping differentially expressed genes from two bulk-seq datasets and an scRNA-seq dataset. A model constructed using these three genes showed good performance in predicting liver metastasis and the survival of CRC patients. Analysis of two bulk-seq datasets and clinical samples by IHC validated the expression of these three genes in CRC tissues.</p> Conclusions <p>Low proportion of activated MCs is associated with liver metastasis in CRC. The model constructed using three MC-enriched genes could effectively predict CRC patients with liver metastasis and their survival outcomes.</p>