Screening Metastatic Biomarkers in Colorectal Cancer Using Machine Learning and Predictive Modeling: In Silico Analysis and Experimental Validation

Colorectal cancer (CRC) is a leading cause of cancer-related mortality, and liver metastasis accounts for the majority of CRC-related deaths. Early detection of metastasis is crucial for improving patient survival but is often delayed due to a lack of noticeable symptoms. Identifying molecular biomarkers associated with CRC metastasis could significantly enhance early detection and therapeutic interventions. In this study, we aimed to discover key metastasis-related biomarkers in CRC using a machine learning (ML) approach, combined with experimental validation, to provide new insights into disease progression and potential treatment strategies. Methods: Gene expression profiles of CRC patients with liver metastasis were obtained from the GSE۴۱۵۶۸ dataset. To identify differentially expressed genes (DEGs) between primary CRC tumors and metastatic liver samples, we conducted a comparative analysis. Feature selection techniques, including LASSO (Least Absolute Shrinkage and Selection Operator) and Penalized-SVM (Support Vector Machine), were applied to pinpoint the most relevant DEGs. Genes consistently selected by both algorithms were further investigated. Experimental validation of selected biomarkers was performed using quantitative real-time PCR (qRT-PCR) on CRC tissue samples to confirm the differential expression patterns observed in silico. Results: The feature selection process using LASSO and Penalized-SVM identified ۱۱ common DEGs associated with CRC metastasis. Of these, seven genes were found to have significant prognostic value in colorectal cancer progression. Notably, the expression of the MMP۳ gene was significantly decreased in stage IV CRC compared to earlier stages (P < ۰.۰۱), while WNT۱۱ expression was significantly elevated in the metastatic stage (P < ۰.۰۰۱). Further analysis revealed that WNT۵a, TNFSF۱۱, and MMP۳ expression levels were lower in liver metastasis samples, whereas WNT۱۱ expression was markedly higher compared to primary tumors. These findings suggest that these genes play crucial roles in the metastatic process and may serve as potential biomarkers for CRC liver metastasis. Conclusion: This study successfully identified a set of potential biomarkers related to CRC liver metastasis through a machine learning-based approach, coupled with experimental validation. The differential expression patterns of MMP۳, WNT۱۱, WNT۵a, and TNFSF۱۱ highlight their potential as diagnostic biomarkers and therapeutic targets for CRC metastasis. These findings provide valuable insights into the molecular mechanisms underlying CRC progression and may contribute to the development of innovative therapeutic strategies for managing metastatic CRC. Future research should focus on validating these biomarkers in larger patient cohorts and exploring their potential clinical applications.

کلیدواژه ها:

Colorectal cancer (CRC) ، Liver metastasis ، Machine learning ، Differentially expressed genes (DEGs) ، Biomarkers