Integrative Analysis of Biomarkers Through Machine Learning Identifies Stemness Features in Colorectal Cancer
Open Access
- 8 September 2021
- journal article
- research article
- Published by Frontiers Media SA in Frontiers in Cell and Developmental Biology
Abstract
Background: Cancer stem cells (CSCs), which are characterized by self-renewal and plasticity, are highly correlated with tumor metastasis and drug resistance. To fully understand the role of CSCs in colorectal cancer (CRC), we evaluated the stemness traits and prognostic value of stemness-related genes in CRC. Methods: In this study, the data from 616 CRC patients from The Cancer Genome Atlas (TCGA) were assessed and subtyped based on the mRNA expression-based stemness index (mRNAsi). The correlations of cancer stemness with the immune microenvironment, tumor mutational burden (TMB), and N6-methyladenosine (m6A) RNA methylation regulators were analyzed. Weighted gene co-expression network analysis (WGCNA) was performed to identify the crucial stemness-related genes and modules. Furthermore, a prognostic expression signature was constructed using the Lasso-penalized Cox regression analysis. The signature was validated via multiplex immunofluorescence staining of tissue samples in an independent cohort of 48 CRC patients. Results: This study suggests that high-mRNAsi scores are associated with poor overall survival in stage IV CRC patients. Moreover, the levels of TMB and m6A RNA methylation regulators were positively correlated with mRNAsi scores, and low-mRNAsi scores were characterized by increased immune activity in CRC. The analysis identified 34 key genes as candidate prognosis biomarkers. Finally, a three-gene prognostic signature (PARPBP, KNSTRN, and KIF2C) was explored together with specific clinical features to construct a nomogram, which was successfully validated in an external cohort. Conclusion: There is a unique correlation between CSCs and the prognosis of CRC patients, and the novel biomarkers related to cell stemness could accurately predict the clinical outcomes of these patients.This publication has 59 references indexed in Scilit:
- ROS Production and NF-κB Activation Triggered by RAC1 Facilitate WNT-Driven Intestinal Stem Cell Proliferation and Colorectal Cancer InitiationCell Stem Cell, 2013
- Mitotic centromere‐associated kinase (MCAK/Kif2C) regulates cellular senescence in human primary cells through a p53‐dependent pathwayFEBS Letters, 2012
- Neoadjuvant Treatment Response As an Early Response Indicator for Patients With Rectal CancerJournal of Clinical Oncology, 2012
- The cBio Cancer Genomics Portal: An Open Platform for Exploring Multidimensional Cancer Genomics DataCancer Discovery, 2012
- clusterProfiler: an R Package for Comparing Biological Themes Among Gene ClustersOMICS: A Journal of Integrative Biology, 2012
- L1 Penalized Estimation in the Cox Proportional Hazards ModelBiometrical Journal, 2010
- Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1Nature, 2009
- Identification and expansion of human colon-cancer-initiating cellsNature, 2006
- A positive feedback loop couples Ras activation and CD44 alternative splicingGenes & Development, 2006
- Macrophages: Obligate Partners for Tumor Cell Migration, Invasion, and MetastasisCell, 2006