Integrative machine learning approach to uncover gene expression and immune landscape insights in colorectal cancer racial disparities 2540
Year: 2025
Type: article
Study design: Observational Study
Abstract: Description Background Racial disparities in colorectal cancer (CRC) are driven by unique tumor biology. African American (AA) and European American (EA) CRC patients exhibit distinct gene expression and tumor microenvironment (TME) profiles. This study leverages gene expression, TME profiling, and machine learning to identify race-specific biomarkers for precision oncology. Methods Differentially expressed genes and TME composition were analyzed in AA and EA CRC patients. Immune cell distributions in the TME were visualized using t-SNE, and machine learning models (XGBoost, Random Forest, SPLSDA) were applied to prioritize biomarkers based on gene expression and TME features. Results AA tumors upregulated IL33, BRAF, and PIK3CA, indicating a pro-inflammatory, proliferative phenotype linked to aggressive CRC, while EA tumors upregulated KRAS, VEGFA, and BCL2, suggesting angiogenic and anti-apoptotic pathways. TME analysis revealed an inflammatory profile in AA tumors, with higher epithelial, B, and myeloid cells, whereas EA tumors had increased plasma and B cells, indicating adaptive immunity. Unique immune patterns by race identified through AI/ML. Race-specific biomarkers included BRAF V600E for both groups, with strategies suggesting immune modulation for AA and angiogenesis inhibition for EA patients. Conclusion The study underscores distinct inflammatory and angiogenic mechanisms in AA and EA CRC, supporting precision oncology approaches to improve outcomes across racial groups. Funding Sources This work was partially supported by the National Institutes of Health (Award Numbers SC1CA193758 and U54CA118638), the National Cancer Institute (Award Number R01CA256724), the American Cancer Society (Award Number CHERC-MSI-21-166-01-CHERC-MSI), Department of Defense (Award Number W81XWH1810429). Topic Categories Computational and Systems Immunology (COMP)
Source: The Journal of Immunology
Institutions: Morehouse School of Medicine, Alabama State University
Language: English
Subfield: Oncology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4416448847
FWCI:
Cites:
Cited by:
Related to:
